> Continuous Process Improvement > Customer Service > Genesys Cloud CX > Data Template

Your Customer Service Data Template

Genesys Cloud CX

Customer Service Attributes (20) Customer Service Activities (13) Extraction Guides (3)

Your Customer Service Data Template

This template provides a clear roadmap for collecting the essential data needed to analyze and optimize your customer service process. It outlines recommended attributes, key activities to track, and practical guidance on extracting this information from your source systems. Use it to build a robust event log for your process mining initiatives.

Recommended attributes to collect
Key activities to track
Extraction guidance

New to event logs? Learn how to create a process mining event log.

Download Template

Customer Service Attributes

These are the recommended data fields to include in your event log for comprehensive customer service analysis.

5 Required 6 Recommended 9 Optional

	Name	Description
	Activity Name ActivityName	The name of the specific event or task that occurred at a point in time within the service request lifecycle.
Description This attribute describes a specific step or status change within the customer service process. Each activity represents a distinct event, such as 'Agent Accepted Interaction' or 'Wrap-Up Code Assigned'. Analyzing the sequence and frequency of these activities is fundamental to process mining. It helps visualize process flows, identify deviations, and pinpoint bottlenecks or inefficient steps. Why it matters Activities form the backbone of the process map, allowing for the visualization and analysis of the actual process flow against the designed flow. Where to get This is typically derived by mapping system events, agent states, or specific audit trail entries in Genesys Cloud CX to standardized activity names. Examples Interaction StartedAgent Accepted InteractionAfter-Call Work EndedService Request Re-Opened
	Service Request ServiceRequest	The primary identifier for a customer service interaction, linking all related activities.
Description The Service Request, often referred to as a Ticket or Case, serves as the primary identifier for a single customer inquiry or issue. It groups all related events, from initial contact to final resolution, into a single process instance. Analyzing by Service Request allows for a complete, end-to-end view of the customer journey and the internal handling process. It is the foundation for calculating key metrics like total resolution time and first contact resolution rate. Why it matters This is the essential Case ID that connects all process steps, enabling a holistic analysis of each customer interaction from start to finish. Where to get This is the conceptual case identifier. In Genesys, this often corresponds to the Conversation ID, but may be a custom identifier depending on the implementation. Examples SR-20240521-00123SR-20240521-00124SR-20240522-00001
	Start Time EventTime	The timestamp indicating when an activity or event started.
Description This attribute provides the precise date and time for when each activity occurred. It is essential for sequencing events correctly and for all time-based analysis. The Start Time is used to construct the process flow in chronological order and is the basis for calculating durations, wait times, and cycle times between different activities. Accurate timestamps are critical for reliable process analysis and performance monitoring. Why it matters This timestamp orders all activities chronologically, making it possible to analyze process flows, durations, and bottlenecks accurately. Where to get Found in event logs or interaction details in Genesys Cloud CX, often associated with every recorded system or user event. Examples 2024-05-21T10:00:15Z2024-05-21T10:02:30Z2024-05-21T10:15:00Z
	Last Data Update LastDataUpdate	The timestamp of the most recent data refresh.
Description This attribute indicates the last time the dataset was updated from the source system. It provides context on the freshness of the data being analyzed, which is critical for making timely and relevant business decisions. Dashboards and reports should display this information prominently so that users are aware of the data's currency and can assess the validity of the insights. Why it matters Indicates data freshness, ensuring that analyses and decisions are based on up-to-date information. Where to get This value is generated and stamped onto the dataset at the time of data extraction or loading into the process mining tool. Examples 2024-05-23T04:00:00Z2024-05-24T04:00:00Z
	Source System SourceSystem	The system from which the data was extracted.
Description This attribute identifies the originating system for the event data, which in this case is Genesys Cloud CX. In environments with multiple integrated systems, this field is crucial for data lineage, troubleshooting, and understanding the context of the data. It helps ensure that analysis is based on the correct data source and allows for filtering or segmenting data when multiple systems contribute to a single process view. Why it matters Identifies the data's origin, which is crucial for data governance, validation, and when merging data from multiple systems. Where to get This is typically a static value added during the data extraction and transformation process to label the origin of the records. Examples Genesys Cloud CXGenesysCloudCX_US1Genesys
	Agent ID AgentId	The unique identifier for the agent who handled the interaction or activity.
Description The Agent ID is a unique key for each customer service representative. This attribute is essential for any analysis related to agent performance, workload distribution, and efficiency. It allows for filtering the process map to see how specific agents or teams handle requests, comparing their performance metrics like resolution time, rework rate, and adherence to standard procedures. This is a cornerstone for the 'Agent Performance & Efficiency' dashboard. Why it matters This attribute connects process activities to specific employees, enabling analysis of individual performance, workload, and team efficiency. Where to get Available in Genesys Cloud CX conversation detail records, associated with the user who handled a specific segment of the interaction. Examples a1b2c3d4-e5f6-7890-1234-567890abcdeff0e9d8c7-b6a5-4321-fedc-ba0987654321
	Agent Name AgentName	The full name of the agent who handled the interaction or activity.
Description The Agent Name provides a human-readable identifier for the service representative, corresponding to the Agent ID. This attribute makes dashboards and reports more intuitive for managers and team leads. It is used extensively in agent-centric analyses to review performance, identify training needs, and ensure fair workload distribution without needing to cross-reference system IDs. Why it matters Provides a user-friendly name for the agent, making it easier to analyze performance and communicate findings without using technical IDs. Where to get Retrieved from the Users or Directory service in Genesys Cloud CX by looking up the AgentId. Examples John SmithJane DoePeter Jones
	Communication Channel MediaType	The communication channel used for the interaction, such as voice, chat, or email.
Description This attribute specifies the medium through which the customer and agent communicated. Common channels include voice, chat, email, and social media. Analyzing performance by channel, as done in the 'Communication Channel Performance' dashboard, helps businesses understand channel-specific volumes, resolution times, and customer satisfaction. This insight is critical for optimizing channel strategy and resource allocation. Why it matters Segmenting the process by communication channel is crucial for understanding channel-specific performance, customer behavior, and resource needs. Where to get A standard field within Genesys Cloud CX conversation detail records. Examples voicechatemailmessage
	End Time EventEndTime	The timestamp indicating when an activity or event ended.
Description This attribute records the precise moment an activity concludes. Paired with the Start Time, it enables the exact calculation of the duration for each individual activity. Analyzing activity durations is key to identifying which steps in the process consume the most time, helping to locate inefficiencies and opportunities for optimization. For example, it can highlight long 'Hold Placed on Interaction' times or extended 'After-Call Work' durations. Why it matters Enables the precise calculation of individual activity durations, which is essential for identifying time-consuming steps and performance bottlenecks. Where to get Found in event logs or interaction details in Genesys Cloud CX. It can also be derived from the start time of the subsequent event. Examples 2024-05-21T10:02:30Z2024-05-21T10:15:00Z2024-05-21T10:18:45Z
	Queue Name QueueName	The name of the queue to which the interaction was routed.
Description This attribute identifies the specific queue that an interaction waited in before being assigned to an agent. Analyzing data by Queue Name is essential for the 'Service Queue Bottleneck Detection' dashboard. It helps managers understand workload distribution across different skill groups or service lines, measure wait times per queue, and identify queues that are consistently understaffed or overloaded. This information is key to optimizing resource allocation and improving customer wait times. Why it matters Helps identify bottlenecks and analyze workload distribution by showing where service requests are waiting for assignment. Where to get Available in Genesys Cloud CX conversation detail records. Each interaction can pass through one or more queues. Examples Tier 1 Support - VoiceBilling Inquiries - ChatTechnical Support - Email
	Wrap-Up Code WrapUpCode	A code assigned by an agent at the end of an interaction to categorize its outcome or topic.
Description The Wrap-Up Code is a label selected by an agent to classify the nature or resolution of a customer interaction. These codes provide structured data about why customers are contacting support. Analyzing Wrap-Up Codes helps identify common issue types, track resolution outcomes, and measure the frequency of specific requests. This data is valuable for root cause analysis and understanding service demand patterns. Why it matters Categorizes the outcome of an interaction, providing structured data for analyzing common issues, resolution effectiveness, and contact reasons. Where to get Available in Genesys Cloud CX conversation detail records, specifically in the session details for agent participants. Examples Password ResetBilling Dispute ResolvedProduct Information RequestEscalated to Tier 2
	Conversation ID ConversationId	The unique identifier assigned by Genesys to an entire conversation.
Description The Conversation ID is the primary technical key in Genesys Cloud CX that groups all related interactions, segments, and participants for a single customer conversation. While the conceptual 'Service Request' is used as the Case ID, the Conversation ID is the underlying key used to retrieve all related data from the Genesys APIs. It is essential for data extraction, joining different data sets, and for technical validation of the process data. Why it matters This is the primary technical key in Genesys, essential for data extraction, troubleshooting, and linking back to the source system. Where to get This is a primary field in all Genesys Cloud CX analytics and conversation APIs. Examples d8a7c6b5-e4f3-2109-8765-fedcba098765c7b6a5d4-f3e2-1098-7654-edcbaf987654
	CSAT Score CustomerSatisfactionScore	The satisfaction score provided by the customer in a post-interaction survey.
Description The Customer Satisfaction Score, or CSAT, is a direct measure of the customer's perception of the service they received. It is typically collected via a survey sent after the interaction is closed. Analyzing CSAT scores in conjunction with process data can reveal how process variations, resolution times, or specific agents impact customer happiness. This is a key outcome metric for evaluating overall process success. Why it matters Directly measures customer happiness, allowing for correlation analysis between process performance and customer outcomes. Where to get This data usually comes from the Genesys Cloud CX Quality Management module or a third-party survey tool integrated with Genesys. Examples 5413
	Is First Contact Resolution IsFirstContactResolution	A flag indicating if the service request was resolved within a single interaction.
Description This boolean attribute identifies cases that were resolved during the first interaction, without requiring any follow-up from the customer or internal transfers. A 'true' value signifies an ideal, efficient resolution. This attribute is the basis for the 'First Contact Resolution Rate' KPI and its corresponding dashboard. Analyzing the characteristics of cases that are not resolved on first contact can reveal opportunities for agent training, knowledge base improvements, or process changes. Why it matters Directly measures efficiency and customer satisfaction, as resolving issues on the first try is a key driver of a positive service experience. Where to get Calculated field, derived by analyzing the event sequence for a case. A case is FCR if it is resolved without certain intervening activities like transfers or re-opens. Examples truefalse
	Is Reopened IsReopened	A flag indicating if a resolved service request was later re-opened.
Description This boolean attribute flags service requests that were set to a resolved or closed state but later became active again. Re-opened cases often indicate that the initial solution was not effective or complete, leading to customer dissatisfaction and additional work. This attribute powers the 'Service Request Re-opening Trends' dashboard and the 'Service Request Re-opening Rate' KPI. Analyzing these cases helps identify the root causes of ineffective resolutions. Why it matters Highlights failures in the resolution process, pointing to issues with solution quality and leading to rework and poor customer experience. Where to get Calculated field, derived by detecting an activity that re-activates the case after a 'Service Request Resolved' activity has occurred. Examples truefalse
	Is SLA Compliant IsSlaCompliant	A flag indicating whether the service request was resolved within its SLA target time.
Description This boolean attribute indicates if a service request met its resolution time target. It is calculated by comparing the 'ServiceResolutionTime' to the 'SlaTargetResolutionTime'. This flag simplifies analysis and visualization for the 'SLA Compliance Overview' dashboard and is the basis for calculating the 'SLA Compliance Rate' KPI. It allows for quick filtering and segmentation of compliant versus non-compliant cases to identify common patterns in SLA breaches. Why it matters Simplifies SLA performance analysis by clearly flagging each case as compliant or breached, enabling root cause analysis of failures. Where to get Calculated field: True if ServiceResolutionTime <= SlaTargetResolutionTime, otherwise False. Examples truefalse
	Processing Time Duration	The time taken to complete a single activity, measured in seconds.
Description This attribute quantifies the duration of an individual process step, calculated as the difference between its End Time and Start Time. Analyzing processing time helps identify which activities are the most time-consuming. This information is vital for performance dashboards, such as the 'Agent Performance & Efficiency' dashboard, to assess agent efficiency and find opportunities for training or process improvements. Why it matters Quantifies the time spent on each activity, helping to pinpoint inefficiencies and measure performance at a granular level. Where to get Calculated field: EventEndTime - EventTime. Some Genesys APIs may provide duration fields directly for certain segments. Examples 12075045
	Service Request Type ServiceRequestType	The classification of the service request, such as 'Inquiry', 'Complaint', or 'Technical Issue'.
Description This attribute categorizes the service request based on its nature or purpose. It allows for segmenting the analysis to understand how different types of requests are handled. For example, the process flow and resolution time for a 'Complaint' may differ significantly from a 'General Inquiry'. This dimension is vital for dashboards like 'Internal Escalation Analysis' to see if certain request types are escalated more often. Why it matters Allows for process segmentation to compare how different types of requests are handled and identify type-specific bottlenecks or inefficiencies. Where to get This information might be captured via an IVR selection, a customer's choice on a web form, or assigned by an agent. In Genesys, it can be stored as a participant attribute or a wrap-up code. Examples Billing InquiryTechnical SupportAccount ManagementProduct Complaint
	Service Resolution Time ServiceResolutionTime	The total time elapsed from the start of the first customer interaction to the final resolution.
Description This metric measures the total duration of a service request, from the moment the customer initiates contact until the issue is marked as resolved. It is a key performance indicator for overall process efficiency and customer experience. This calculated attribute is the primary focus of the 'Service Request Resolution Time' dashboard and the 'Average Service Resolution Time' KPI. Analyzing its distribution helps identify long-running cases and systemic delays. Why it matters This is a critical KPI for measuring overall process efficiency and its impact on the customer experience. Where to get Calculated field: Timestamp of the final resolution activity minus the timestamp of the first customer contact activity. Examples 90010800172800
	SLA Target Resolution Time SlaTargetResolutionTime	The contractually agreed upon target time for resolving the service request.
Description This attribute defines the maximum time allowed for a service request to be resolved, according to the Service Level Agreement (SLA). It serves as the benchmark against which actual resolution times are measured. This data is essential for the 'SLA Compliance Overview' dashboard and the 'SLA Compliance Rate' KPI, enabling the organization to monitor its performance against customer commitments and identify potential breaches before they happen. Why it matters Provides the benchmark for measuring SLA compliance, a critical indicator of service performance and contractual obligation. Where to get This may be stored as a custom attribute on the conversation or derived based on rules involving the queue, customer type, or request type. Examples 86400144003600

Required Recommended Optional

Customer Service Activities

These are the key process steps and milestones to capture in your event log for accurate process discovery.

5 Recommended 8 Optional

	Activity	Description
	Agent Accepted Interaction	Marks the point where an agent accepts the offered interaction and is connected to the customer. This is a key milestone where the direct handling of the service request begins.
Why it matters This activity is crucial for measuring First Contact Resolution and agent handle time. It signifies the start of the agent's work on the request. Where to get Available in the conversation detail records. It is marked by the agent participant's state changing to 'connected', with an associated timestamp. Capture Logged when an agent participant's state changes to 'connected'. Event type explicit
	Interaction Disconnected	This activity signifies the end of the conversation, when all participants have disconnected. This often serves as the technical closure of the service request interaction.
Why it matters This is a definitive end event for the interaction itself. It is crucial for calculating total interaction duration and is often used as a proxy for request closure. Where to get This is captured from the conversation detail records. It corresponds to the end time of the conversation object, typically found in the conversationEnd timestamp. Capture Logged when all participants have disconnected and the conversation ends. Event type explicit
	Interaction Started	This activity marks the beginning of a customer service interaction, such as an inbound call, chat, or email. Genesys Cloud CX explicitly logs this event when a new conversation object is created in the system.
Why it matters This is the primary start event for the service request process. It is essential for calculating the total resolution time and understanding incoming customer demand patterns. Where to get This event is captured from the conversation detail records. It corresponds to the start time of the conversation object itself, typically found in the conversationStart timestamp. Capture Logged when a new conversation is initiated in the system. Event type explicit
	Interaction Transferred	Represents an agent transferring an interaction to another queue or agent. This can be a blind transfer where the agent immediately disconnects or a consult transfer where they speak with the recipient first.
Why it matters This activity is essential for the Internal Escalation Rate KPI. High transfer rates can indicate training needs, incorrect routing, or process complexity. Where to get This is identified in the conversation detail records when a new ACD or agent participant is added to the conversation following the initial agent's involvement, often initiated by a 'transfer' event. Capture Identified by a transfer event in the participant's session data. Event type explicit
	Wrap-Up Code Assigned	An agent assigns a predefined wrap-up code, or disposition code, to the interaction. This explicitly categorizes the outcome of the service request, such as 'Resolved' or 'Escalated'.
Why it matters Wrap-up codes are a primary source for determining the business outcome of a request. They are vital for calculating resolution rates and segmenting cases for analysis. Where to get This is an explicit event found in the conversation detail records, typically within the agent participant's session data. The wrapUp code and timestamp are recorded. Capture Logged when an agent selects a wrap-up code for the interaction. Event type explicit
	After-Call Work Ended	This marks the end of the after-call work period for an agent. At this point, the agent becomes available to handle another interaction.
Why it matters This activity, combined with 'After-Call Work Started', provides the precise duration of the wrap-up phase, helping to analyze agent productivity and process overhead. Where to get This is inferred when the agent's state changes from 'acw' to an available state, such as 'idle'. The timestamp of this state change is used. Capture Inferred from the agent's state changing from 'acw' to an available state. Event type inferred
	After-Call Work Started	This activity marks the beginning of the after-call work (ACW) period. The agent has disconnected from the customer but is now in a dedicated state to complete tasks like logging notes or updating systems.
Why it matters Measuring ACW duration is important for understanding agent efficiency and overall handle time. Extended ACW can be a sign of inefficient post-interaction processes. Where to get Captured from the agent participant's session data. It is logged when the agent's state changes to 'acw' after the customer participant disconnects. Capture Logged when an agent participant's state changes to 'acw'. Event type explicit
	Agent Offered Interaction	This event occurs when the system offers an interaction to a specific agent. The agent's status changes to 'alerting' as the system waits for them to accept or reject the conversation.
Why it matters This activity helps analyze agent responsiveness and the effectiveness of the routing algorithm. Delays after this point can indicate agents are not accepting work promptly. Where to get Found in conversation detail records within the participant data for the agent. The participant's session will show a metric with an 'alerting' state and a corresponding timestamp. Capture Logged when the routing engine alerts an agent of a new interaction. Event type explicit
	Conversation Routed to Queue	Represents the moment a new interaction is placed into a specific queue to wait for an available agent. This is an explicit event logged by the Genesys routing engine (ACD).
Why it matters Tracking this activity is critical for measuring queue wait times and identifying bottlenecks in the routing process. High durations between this and agent assignment indicate staffing or routing logic issues. Where to get Captured from conversation detail records, specifically looking at the 'purpose' and 'state' of the ACD participant within the conversation. The 'enterTime' for the queue in the metrics array indicates this event. Capture Logged when the ACD routing engine places a conversation into a queue. Event type explicit
	Customer Satisfaction Survey Sent	This activity occurs when a customer satisfaction (CSAT) survey is sent following the closure of an interaction. This is often triggered automatically by the system based on predefined rules.
Why it matters This activity is necessary to measure the CSAT Survey Delivery Timeliness KPI. Ensuring timely survey delivery helps capture accurate feedback while the experience is fresh in the customer's mind. Where to get This information is typically available from survey or quality management data within Genesys Cloud CX, which can be linked back to the original conversation ID. Capture Logged by the survey module when a survey is dispatched for a specific conversation. Event type explicit
	Hold Placed on Interaction	This activity is recorded when an agent places a customer on hold during an interaction. This is an explicit state change for the agent participant within the conversation.
Why it matters Analyzing hold frequency and duration can reveal process inefficiencies, such as agents frequently needing to search for information or consult with others. Where to get Captured from the agent participant's session data in the conversation detail records. The state changes to 'held' and a timestamp is recorded. Capture Logged when an agent participant's state changes to 'held'. Event type explicit
	Hold Removed from Interaction	Occurs when an agent takes a customer off hold and resumes the conversation. This event marks the end of a hold period and is logged as a state change.
Why it matters Paired with 'Hold Placed on Interaction', this activity allows for precise calculation of total hold time, a key component of overall handle time and customer experience. Where to get Derived from the agent participant's session data when their state changes from 'held' back to 'connected'. Capture Logged when an agent participant's state changes from 'held' to 'connected'. Event type explicit
	Service Request Re-Opened	Represents a situation where a customer contacts the service center again about the same issue shortly after it was considered resolved. This is not an explicit event but is calculated based on business logic.
Why it matters Tracking re-opened requests is key to measuring the Service Request Re-opening Rate. It indicates that the initial resolution was not effective, impacting customer satisfaction and operational efficiency. Where to get This activity is inferred by identifying a new 'Interaction Started' event for the same customer (Customer ID) and related issue (e.g., Service Request Type) within a predefined time window after a previous interaction was resolved. Capture Calculated by linking a new interaction to a recently closed one for the same customer and issue. Event type calculated

Recommended Optional

Extraction Guides

How to get your data from Genesys Cloud CX

Steps

Navigate to the Analytics Workspace: Log in to Genesys Cloud CX. Go to the menu and select Admin, then under Analytics, choose Workspace.
Create a New View: Click the plus icon to add a new tab and select "Create a new view". Choose "Interactions" as the view type. This will open the interaction view editor.
Define the Date Range: In the filter panel on the left, set the desired date and time interval for the data you want to extract. It is recommended to start with a smaller date range, such as one week, to test the process.
Select Columns for the Export: This is the most critical step for data preparation. Click on the "Columns" button. From the available fields, select and add the following columns to your view. These fields contain the necessary data for building the event log:
- Conversation ID
- Conversation Start
- Conversation End
- Media Type
- Participants (This is a complex field containing nested agent, queue, and customer data)
- Metrics.tAnswered
- Metrics.tACW
- Metrics.tHeld
- Surveys
Apply Additional Filters: Use the filter panel to narrow down the data if necessary. You can filter by Media Type (e.g., voice, chat), Queue, Direction (inbound/outbound), or other conversation properties to focus on the specific service request process.
Save and Run the View: Save the view with a descriptive name, for example, "Process Mining Service Request Export". The view will then load the data matching your criteria.
Initiate the Export: Once the data is loaded in your view, click the "Export" button, usually represented by a download icon. Select "Export as CSV" or "Export as JSON". CSV is often easier for initial processing.
Download the Exported File: The export job will run in the background. You can monitor its progress by navigating to Performance > My Exports. Once completed, download the file to your local system.
Transform the Raw Data: The exported file has one row per conversation, with participant and segment data nested within a single column. You must run a transformation script, like the Python script provided in the Query section, to un-nest this data. This script will parse the raw export and create a new file where each row represents a single activity, such as "Agent Accepted Interaction" or "Hold Placed on Interaction".
Implement Custom Logic: As part of the transformation, implement any business-specific logic. For example, the "Service Request Re-Opened" activity must be calculated by analyzing subsequent interactions from the same customer within a defined timeframe.
Finalize the Event Log: The transformation script should output a CSV file with columns matching the required format for ProcessMind, including ServiceRequest, ActivityName, EventTime, and other attributes. Ensure all timestamps are in a consistent format.
Upload to ProcessMind: The final, transformed CSV file is now ready to be uploaded to your process mining tool.

Configuration

Permissions: Your Genesys Cloud CX user role must have the Analytics > Conversation > View permission to access the Analytics Workspace and create interaction views.
Date Range: For the initial extraction, use a representative period of 3-6 months. For very high-volume contact centers, it is recommended to export data in smaller chunks, such as weekly or monthly, to avoid hitting system limits and to manage file sizes.
Key Filters:
- Media Type: Filter for the channels relevant to your customer service process, such as voice, chat, or email.
- Queue: If the process is handled by specific teams, filter by the relevant Queue IDs to ensure you are only extracting data for the desired process.
- Direction: Typically, you will want to filter for inbound interactions to capture customer-initiated service requests.
Export Columns: The Participants column is essential as it contains the detailed timeline of events for agents and queues. Without this column, you cannot extract individual activities. Ensure Conversation ID, Conversation Start, and Conversation End are also included.
Performance Considerations: Exporting a large number of columns or a very long date range can cause the export job to take a significant amount of time. If the job fails or times out, reduce the date range and try again. Genesys Cloud has a limit of approximately 1 million conversations per export job.

a Sample Query config

import pandas as pd
import json
from datetime import datetime, timezone

# This script transforms a Genesys Cloud Conversation Detail Export into a process mining event log.
# It assumes the raw data is in a CSV file named 'genesys_export.csv'.

def transform_genesys_export(file_path):
    # Load the raw export file
    try:
        df = pd.read_csv(file_path)
    except FileNotFoundError:
        print(f"Error: The file {file_path} was not found.")
        return None

    event_log = []
    source_system = 'Genesys Cloud CX'
    last_data_update = datetime.now(timezone.utc).isoformat()

    # Rename columns for easier access
    df.rename(columns={
        'Conversation ID': 'conversationId',
        'Conversation Start': 'conversationStart',
        'Conversation End': 'conversationEnd',
        'Media Type': 'mediaType'
    }, inplace=True)

    for index, row in df.iterrows():
        service_request_id = row['conversationId']
        media_type = row['mediaType']
        
        # Activity: Interaction Started
        event_log.append({
            'ServiceRequest': service_request_id,
            'ActivityName': 'Interaction Started',
            'EventTime': row['conversationStart'],
            'EventEndTime': None,
            'AgentId': None, 'AgentName': None, 'QueueName': None, 'WrapUpCode': None,
            'MediaType': media_type, 'SourceSystem': source_system, 'LastDataUpdate': last_data_update
        })

        # The 'Participants' column contains a JSON string that needs to be parsed
        participants = json.loads(row['Participants'])

        for p in participants:
            participant_id = p.get('participantId')
            user_id = p.get('userId')
            queue_name = p.get('queueName')
            purpose = p.get('purpose')

            for s in p.get('sessions', []):
                wrap_up_code = s.get('wrapUpCode')
                for seg in s.get('segments', []):
                    seg_start = seg.get('segmentStart')
                    seg_end = seg.get('segmentEnd')
                    seg_type = seg.get('segmentType')

                    activity_name = None
                    if purpose == 'acd' and seg_type == 'interact':
                        activity_name = 'Conversation Routed to Queue'
                    elif purpose == 'agent' and seg_type == 'alert':
                        activity_name = 'Agent Offered Interaction'
                    elif purpose == 'agent' and seg_type == 'interacting':
                        activity_name = 'Agent Accepted Interaction'
                    elif purpose == 'agent' and seg_type == 'hold':
                        activity_name = 'Hold Placed on Interaction'
                        # Also create the Hold Removed event
                        event_log.append({
                            'ServiceRequest': service_request_id,
                            'ActivityName': 'Hold Removed from Interaction',
                            'EventTime': seg_end,
                            'EventEndTime': None,
                            'AgentId': user_id, 'AgentName': '[Configure based on your system]', 'QueueName': queue_name, 'WrapUpCode': None,
                            'MediaType': media_type, 'SourceSystem': source_system, 'LastDataUpdate': last_data_update
                        })
                    elif purpose == 'agent' and seg_type == 'acw':
                        activity_name = 'After-Call Work Started'
                        # Also create ACW Ended and Wrap-Up Assigned events
                        if wrap_up_code:
                            event_log.append({
                                'ServiceRequest': service_request_id,
                                'ActivityName': 'Wrap-Up Code Assigned',
                                'EventTime': seg_end, # Assume code assigned at end of ACW
                                'EventEndTime': None,
                                'AgentId': user_id, 'AgentName': '[Configure based on your system]', 'QueueName': queue_name, 'WrapUpCode': wrap_up_code,
                                'MediaType': media_type, 'SourceSystem': source_system, 'LastDataUpdate': last_data_update
                            })
                        event_log.append({
                            'ServiceRequest': service_request_id,
                            'ActivityName': 'After-Call Work Ended',
                            'EventTime': seg_end,
                            'EventEndTime': None,
                            'AgentId': user_id, 'AgentName': '[Configure based on your system]', 'QueueName': queue_name, 'WrapUpCode': wrap_up_code,
                            'MediaType': media_type, 'SourceSystem': source_system, 'LastDataUpdate': last_data_update
                        })
                    
                    if seg.get('disconnectType') == 'transfer':
                        activity_name = 'Interaction Transferred'

                    if activity_name:
                        event_log.append({
                            'ServiceRequest': service_request_id,
                            'ActivityName': activity_name,
                            'EventTime': seg_start,
                            'EventEndTime': seg_end,
                            'AgentId': user_id, 'AgentName': '[Configure based on your system]', 'QueueName': queue_name, 'WrapUpCode': None,
                            'MediaType': media_type, 'SourceSystem': source_system, 'LastDataUpdate': last_data_update
                        })
        
        # Activity: Interaction Disconnected
        event_log.append({
            'ServiceRequest': service_request_id,
            'ActivityName': 'Interaction Disconnected',
            'EventTime': row['conversationEnd'],
            'EventEndTime': None,
            'AgentId': None, 'AgentName': None, 'QueueName': None, 'WrapUpCode': None,
            'MediaType': media_type, 'SourceSystem': source_system, 'LastDataUpdate': last_data_update
        })

        # Activity: Customer Satisfaction Survey Sent (from surveys column)
        if pd.notna(row.get('Surveys')):
            surveys = json.loads(row['Surveys'])
            for survey in surveys:
                if survey.get('surveyStatus') == 'Sent':
                    # Survey sent time is not always directly available, using conversation end as an approximation
                    event_log.append({
                        'ServiceRequest': service_request_id,
                        'ActivityName': 'Customer Satisfaction Survey Sent',
                        'EventTime': survey.get('surveySentDate', row['conversationEnd']),
                        'EventEndTime': None,
                        'AgentId': None, 'AgentName': None, 'QueueName': None, 'WrapUpCode': None,
                        'MediaType': media_type, 'SourceSystem': source_system, 'LastDataUpdate': last_data_update
                    })

    # NOTE: 'Service Request Re-Opened' requires custom cross-conversation logic.
    # This logic would need to be implemented here, analyzing the entire dataset
    # to find subsequent interactions from the same customer within a specific timeframe.

    final_df = pd.DataFrame(event_log)
    # Sort by case and event time to create a chronological log
    final_df.sort_values(by=['ServiceRequest', 'EventTime'], inplace=True)
    
    return final_df

if __name__ == '__main__':
    # Specify the path to your downloaded Genesys export CSV file
    input_file = 'genesys_export.csv'
    output_file = 'genesys_event_log.csv'
    
    event_log_df = transform_genesys_export(input_file)
    
    if event_log_df is not None:
        event_log_df.to_csv(output_file, index=False)
        print(f"Transformation complete. Event log saved to {output_file}")

Steps

Prerequisites and Setup: Ensure you have Python 3.8+ installed along with the pandas and purecloud-platform-client-v2 libraries. You can install them using pip: pip install pandas purecloud-platform-client-v2.
Create Genesys Cloud OAuth Client: Log in to your Genesys Cloud CX admin account. Navigate to Integrations > OAuth and create a new client with the 'Client Credentials' grant type. Assign a role that includes the analytics:conversationDetail:view permission.
Configure Script Credentials: Save the Client ID and Client Secret from the previous step. In the Python script provided below, replace the placeholder values for CLIENT_ID, CLIENT_SECRET, and GENESYS_CLOUD_REGION with your specific credentials.
Define the Extraction Interval: Locate the query_body variable in the script. Modify the interval value to specify the date and time range for the data you wish to extract. It is recommended to start with a small range, like a single day, for testing.
Execute the Python Script: Run the script from your terminal using python your_script_name.py. The script will authenticate with the Genesys Cloud API, submit an asynchronous analytics job, and print the job ID.
Monitor Job Execution: The script will automatically poll the API every 15 seconds to check the status of the job. You will see status updates printed to the console. This step can take several minutes or longer, depending on the volume of data requested.
Process and Transform Data: Once the job is complete, the script downloads the resulting JSON data. It then iterates through each conversation record, parsing the participant, session, and segment data to identify and structure the key activities and attributes.
Handle Calculated Events: The script includes a post-processing step to identify 'Service Request Re-Opened' events. This logic, which you can customize, checks for new interactions from the same customer within a specified time window after a previous interaction was resolved.
Generate the Event Log: The transformed data is compiled into a pandas DataFrame, which mirrors the structure of a process mining event log.
Export the Final CSV File: The script saves the final DataFrame as a CSV file named genesys_event_log.csv. This file is formatted for direct upload into the ProcessMind platform.

Configuration

API Credentials: An OAuth Client ID and Client Secret are required for authentication. These must be created within the Genesys Cloud CX admin interface.
API Permissions: The associated OAuth client role must have the analytics:conversationDetail:view permission to access the necessary conversation data.
Extraction Interval: The interval parameter in the API query body controls the date range of the extraction. For large environments, it is best practice to run the extraction in smaller, sequential batches (e.g., one week at a time) to avoid API timeouts and manage memory usage.
Data Filters: The API query body can be modified to include conversationFilters, segmentFilters, or evaluationFilters to narrow the scope of the extraction. For example, you could filter by specific queueId or mediaType to target a particular part of the customer service process.
Re-Open Logic: The time window for identifying a 'Service Request Re-Opened' event is configurable within the script. The default is set to 72 hours but should be adjusted to match your organization's business rules.
Environment: A local Python environment with the required SDK and libraries is necessary to run the script. No direct access to the Genesys Cloud CX server or database is needed.

a Sample Query python

import PureCloudPlatformClientV2
import pandas as pd
import requests
import json
import time
import os
from datetime import datetime, timedelta

class=class="hl-string">"hl-comment"># --- Configuration ---
class=class="hl-string">"hl-comment"># Replace with your Genesys Cloud credentials and region
CLIENT_ID = class="hl-string">"[Your_Client_Id]"
CLIENT_SECRET = class="hl-string">"[Your_Client_Secret]"
GENESYS_CLOUD_REGION = class="hl-string">"[Your_Genesys_Cloud_Region]" class=class="hl-string">"hl-comment"># e.g., mypurecloud.com, mypurecloud.de

class=class="hl-string">"hl-comment"># Set the API environment
PureCloudPlatformClientV2.configuration.host = fclass="hl-string">'https://api.{GENESYS_CLOUD_REGION}'

class=class="hl-string">"hl-comment"># --- Authentication ---
api_client = PureCloudPlatformClientV2.api_client.ApiClient().get_client_credentials_token(CLIENT_ID, CLIENT_SECRET)
analytics_api = PureCloudPlatformClientV2.AnalyticsApi(api_client)


def parse_conversation(conv_detail):
    class="hl-string">"""Parses a single conversation detail object into multiple event log rows."""
    events = []
    conversation_id = conv_detail.get(class="hl-string">'conversationId')
    source_system = class="hl-string">'Genesys Cloud CX'
    last_data_update = datetime.utcnow().isoformat() + class="hl-string">'Z'

    class=class="hl-string">"hl-comment"># Event: Interaction Started
    events.append({
        class="hl-string">'ServiceRequest': conversation_id,
        class="hl-string">'ActivityName': class="hl-string">'Interaction Started',
        class="hl-string">'EventTime': conv_detail.get(class="hl-string">'conversationStart'),
        class="hl-string">'EventEndTime': None,
        class="hl-string">'AgentId': None, class="hl-string">'AgentName': None, class="hl-string">'QueueName': None, class="hl-string">'WrapUpCode': None,
        class="hl-string">'MediaType': next((p.get(class="hl-string">'sessions')[0].get(class="hl-string">'mediaType') for p in conv_detail.get(class="hl-string">'participants', []) if p.get(class="hl-string">'purpose') == class="hl-string">'customer'), None),
        class="hl-string">'SourceSystem': source_system, class="hl-string">'LastDataUpdate': last_data_update
    })

    for p in conv_detail.get(class="hl-string">'participants', []):
        agent_id = p.get(class="hl-string">'userId')
        agent_name = p.get(class="hl-string">'participantName')

        for s in p.get(class="hl-string">'sessions', []):
            media_type = s.get(class="hl-string">'mediaType')

            for seg in s.get(class="hl-string">'segments', []):
                seg_start = seg.get(class="hl-string">'segmentStart')
                seg_end = seg.get(class="hl-string">'segmentEnd')

                if seg.get(class="hl-string">'segmentType') == class="hl-string">'alert':
                    events.append({
                        class="hl-string">'ServiceRequest': conversation_id, class="hl-string">'ActivityName': class="hl-string">'Agent Offered Interaction', class="hl-string">'EventTime': seg_start, class="hl-string">'EventEndTime': seg_end,
                        class="hl-string">'AgentId': agent_id, class="hl-string">'AgentName': agent_name, class="hl-string">'QueueName': s.get(class="hl-string">'queueName'), class="hl-string">'WrapUpCode': None, class="hl-string">'MediaType': media_type,
                        class="hl-string">'SourceSystem': source_system, class="hl-string">'LastDataUpdate': last_data_update
                    })

                if seg.get(class="hl-string">'segmentType') == class="hl-string">'interact' and p.get(class="hl-string">'purpose') == class="hl-string">'agent':
                    events.append({
                        class="hl-string">'ServiceRequest': conversation_id, class="hl-string">'ActivityName': class="hl-string">'Agent Accepted Interaction', class="hl-string">'EventTime': seg_start, class="hl-string">'EventEndTime': seg_end,
                        class="hl-string">'AgentId': agent_id, class="hl-string">'AgentName': agent_name, class="hl-string">'QueueName': s.get(class="hl-string">'queueName'), class="hl-string">'WrapUpCode': None, class="hl-string">'MediaType': media_type,
                        class="hl-string">'SourceSystem': source_system, class="hl-string">'LastDataUpdate': last_data_update
                    })

                if seg.get(class="hl-string">'segmentType') == class="hl-string">'hold':
                    events.append({
                        class="hl-string">'ServiceRequest': conversation_id, class="hl-string">'ActivityName': class="hl-string">'Hold Placed on Interaction', class="hl-string">'EventTime': seg_start, class="hl-string">'EventEndTime': None,
                        class="hl-string">'AgentId': agent_id, class="hl-string">'AgentName': agent_name, class="hl-string">'QueueName': s.get(class="hl-string">'queueName'), class="hl-string">'WrapUpCode': None, class="hl-string">'MediaType': media_type,
                        class="hl-string">'SourceSystem': source_system, class="hl-string">'LastDataUpdate': last_data_update
                    })
                    if seg_end:
                        events.append({
                            class="hl-string">'ServiceRequest': conversation_id, class="hl-string">'ActivityName': class="hl-string">'Hold Removed from Interaction', class="hl-string">'EventTime': seg_end, class="hl-string">'EventEndTime': None,
                            class="hl-string">'AgentId': agent_id, class="hl-string">'AgentName': agent_name, class="hl-string">'QueueName': s.get(class="hl-string">'queueName'), class="hl-string">'WrapUpCode': None, class="hl-string">'MediaType': media_type,
                            class="hl-string">'SourceSystem': source_system, class="hl-string">'LastDataUpdate': last_data_update
                        })

                if seg.get(class="hl-string">'segmentType') == class="hl-string">'acw':
                    events.append({
                        class="hl-string">'ServiceRequest': conversation_id, class="hl-string">'ActivityName': class="hl-string">'After-Call Work Started', class="hl-string">'EventTime': seg_start, class="hl-string">'EventEndTime': seg_end,
                        class="hl-string">'AgentId': agent_id, class="hl-string">'AgentName': agent_name, class="hl-string">'QueueName': s.get(class="hl-string">'queueName'), class="hl-string">'WrapUpCode': p.get(class="hl-string">'wrapUp', {}).get(class="hl-string">'code'), class="hl-string">'MediaType': media_type,
                        class="hl-string">'SourceSystem': source_system, class="hl-string">'LastDataUpdate': last_data_update
                    })
                    if seg_end:
                         events.append({
                            class="hl-string">'ServiceRequest': conversation_id, class="hl-string">'ActivityName': class="hl-string">'After-Call Work Ended', class="hl-string">'EventTime': seg_end, class="hl-string">'EventEndTime': None,
                            class="hl-string">'AgentId': agent_id, class="hl-string">'AgentName': agent_name, class="hl-string">'QueueName': s.get(class="hl-string">'queueName'), class="hl-string">'WrapUpCode': p.get(class="hl-string">'wrapUp', {}).get(class="hl-string">'code'), class="hl-string">'MediaType': media_type,
                            class="hl-string">'SourceSystem': source_system, class="hl-string">'LastDataUpdate': last_data_update
                        })
                
                if seg.get(class="hl-string">'segmentType') == class="hl-string">'transmitting' and seg.get(class="hl-string">'disconnectType') == class="hl-string">'transfer':
                    events.append({
                        class="hl-string">'ServiceRequest': conversation_id, class="hl-string">'ActivityName': class="hl-string">'Interaction Transferred', class="hl-string">'EventTime': seg_start, class="hl-string">'EventEndTime': None,
                        class="hl-string">'AgentId': agent_id, class="hl-string">'AgentName': agent_name, class="hl-string">'QueueName': s.get(class="hl-string">'queueName'), class="hl-string">'WrapUpCode': None, class="hl-string">'MediaType': media_type,
                        class="hl-string">'SourceSystem': source_system, class="hl-string">'LastDataUpdate': last_data_update
                    })
            
            if s.get(class="hl-string">'flow', {}).get(class="hl-string">'outcomeValue') == class="hl-string">'SURVEY_OFFERED':
                events.append({
                    class="hl-string">'ServiceRequest': conversation_id, class="hl-string">'ActivityName': class="hl-string">'Customer Satisfaction Survey Sent', class="hl-string">'EventTime': s.get(class="hl-string">'flow', {}).get(class="hl-string">'flowEnd'), class="hl-string">'EventEndTime': None,
                    class="hl-string">'AgentId': None, class="hl-string">'AgentName': None, class="hl-string">'QueueName': None, class="hl-string">'WrapUpCode': None, class="hl-string">'MediaType': media_type,
                    class="hl-string">'SourceSystem': source_system, class="hl-string">'LastDataUpdate': last_data_update
                })

        if p.get(class="hl-string">'purpose') == class="hl-string">'acd':
             for s in p.get(class="hl-string">'sessions', []):
                for seg in s.get(class="hl-string">'segments', []):
                    if seg.get(class="hl-string">'segmentType') == class="hl-string">'interact':
                         events.append({
                            class="hl-string">'ServiceRequest': conversation_id, class="hl-string">'ActivityName': class="hl-string">'Conversation Routed to Queue', class="hl-string">'EventTime': seg.get(class="hl-string">'segmentStart'), class="hl-string">'EventEndTime': seg.get(class="hl-string">'segmentEnd'),
                            class="hl-string">'AgentId': None, class="hl-string">'AgentName': None, class="hl-string">'QueueName': s.get(class="hl-string">'queueName'), class="hl-string">'WrapUpCode': None, class="hl-string">'MediaType': s.get(class="hl-string">'mediaType'),
                            class="hl-string">'SourceSystem': source_system, class="hl-string">'LastDataUpdate': last_data_update
                        })

        if p.get(class="hl-string">'wrapUp'):
            events.append({
                class="hl-string">'ServiceRequest': conversation_id, class="hl-string">'ActivityName': class="hl-string">'Wrap-Up Code Assigned', class="hl-string">'EventTime': p[class="hl-string">'wrapUp'].get(class="hl-string">'wrapUpEnd'), class="hl-string">'EventEndTime': None,
                class="hl-string">'AgentId': agent_id, class="hl-string">'AgentName': agent_name, class="hl-string">'QueueName': s.get(class="hl-string">'queueName'), class="hl-string">'WrapUpCode': p[class="hl-string">'wrapUp'].get(class="hl-string">'code'), class="hl-string">'MediaType': s.get(class="hl-string">'mediaType'),
                class="hl-string">'SourceSystem': source_system, class="hl-string">'LastDataUpdate': last_data_update
            })

    class=class="hl-string">"hl-comment"># Event: Interaction Disconnected
    events.append({
        class="hl-string">'ServiceRequest': conversation_id, class="hl-string">'ActivityName': class="hl-string">'Interaction Disconnected', class="hl-string">'EventTime': conv_detail.get(class="hl-string">'conversationEnd'), class="hl-string">'EventEndTime': None,
        class="hl-string">'AgentId': None, class="hl-string">'AgentName': None, class="hl-string">'QueueName': None, class="hl-string">'WrapUpCode': None,
        class="hl-string">'MediaType': next((p.get(class="hl-string">'sessions')[0].get(class="hl-string">'mediaType') for p in conv_detail.get(class="hl-string">'participants', []) if p.get(class="hl-string">'purpose') == class="hl-string">'customer'), None),
        class="hl-string">'SourceSystem': source_system, class="hl-string">'LastDataUpdate': last_data_update
    })
    
    return events


def main():
    class=class="hl-string">"hl-comment"># --- Define Analytics Query ---
    class=class="hl-string">"hl-comment"># Define your time interval here. Format: YYYY-MM-DDTHH:MM:SS.sssZ/YYYY-MM-DDTHH:MM:SS.sssZ
    interval = class="hl-string">"2023-10-01T00:00:00.000Z/2023-10-02T00:00:00.000Z"
    
    query_body = {
        class="hl-string">"interval": interval,
        class="hl-string">"paging": {
            class="hl-string">"pageSize": 100,
            class="hl-string">"pageNumber": 1
        }
    }

    class=class="hl-string">"hl-comment"># --- Submit and Monitor Job ---
    try:
        print(fclass="hl-string">"Submitting analytics job for interval: {interval}")
        api_response = analytics_api.post_analytics_conversations_details_jobs(query_body)
        job_id = api_response.job_id
        print(fclass="hl-string">"Job submitted successfully. Job ID: {job_id}")

        while True:
            job_status = analytics_api.get_analytics_conversations_details_job(job_id)
            print(fclass="hl-string">"Job status: {job_status.state}. Progress: {job_status.percent_complete}%")
            if job_status.state == class="hl-string">'FULFILLED':
                print(class="hl-string">"Job completed successfully.")
                break
            elif job_status.state in [class="hl-string">'FAILED', class="hl-string">'CANCELLED']:
                print(fclass="hl-string">"Job failed or was cancelled. State: {job_status.state}, Error: {job_status.error}")
                return
            time.sleep(15)

        class=class="hl-string">"hl-comment"># --- Fetch and Process Results ---
        all_events = []
        results_response = analytics_api.get_analytics_conversations_details_job_results(job_id)
        
        if not results_response.conversations:
            print(class="hl-string">"No conversations found for the given interval.")
            return
            
        print(fclass="hl-string">"Processing {len(results_response.conversations)} conversations...")
        for conv in results_response.conversations:
            all_events.extend(parse_conversation(conv))

        df = pd.DataFrame(all_events)
        df[class="hl-string">'EventTime'] = pd.to_datetime(df[class="hl-string">'EventTime'], errors=class="hl-string">'coerce')
        df = df.dropna(subset=[class="hl-string">'EventTime'])
        df = df.sort_values(by=[class="hl-string">'ServiceRequest', class="hl-string">'EventTime']).reset_index(drop=True)

        class=class="hl-string">"hl-comment"># --- Post-Processing for class="hl-string">'Service Request Re-Opened' ---
        class=class="hl-string">"hl-comment"># This logic is an example. Adjust the window and criteria to fit your business rules.
        print(class="hl-string">"Applying post-processing for 'Re-Opened' events...")
        reopen_events = []
        df_resolved = df[df[class="hl-string">'WrapUpCode'].str.contains(class="hl-string">'Resolved', na=False)].copy()
        df_resolved = df_resolved.sort_values(class="hl-string">'EventTime').drop_duplicates(class="hl-string">'ServiceRequest', keep=class="hl-string">'last')
        reopen_window = timedelta(hours=72)

        for index, row in df_resolved.iterrows():
            customer_id = next((p.get(class="hl-string">'participantName') for c in results_response.conversations if c.get(class="hl-string">'conversationId') == row[class="hl-string">'ServiceRequest'] for p in c.get(class="hl-string">'participants', []) if p.get(class="hl-string">'purpose') == class="hl-string">'customer'), None)
            if not customer_id: continue
            
            resolved_time = row[class="hl-string">'EventTime']
            future_interactions = df[(df[class="hl-string">'EventTime'] > resolved_time) & (df[class="hl-string">'EventTime'] <= resolved_time + reopen_window)]
            
            for _, future_row in future_interactions.iterrows():
                future_customer_id = next((p.get(class="hl-string">'participantName') for c in results_response.conversations if c.get(class="hl-string">'conversationId') == future_row[class="hl-string">'ServiceRequest'] for p in c.get(class="hl-string">'participants', []) if p.get(class="hl-string">'purpose') == class="hl-string">'customer'), None)
                if customer_id == future_customer_id and future_row[class="hl-string">'ActivityName'] == class="hl-string">'Interaction Started':
                    reopen_events.append({
                        class="hl-string">'ServiceRequest': row[class="hl-string">'ServiceRequest'], class="hl-string">'ActivityName': class="hl-string">'Service Request Re-Opened', class="hl-string">'EventTime': future_row[class="hl-string">'EventTime'],
                        class="hl-string">'SourceSystem': class="hl-string">'Genesys Cloud CX', class="hl-string">'LastDataUpdate': datetime.utcnow().isoformat() + class="hl-string">'Z'
                    })
                    break

        if reopen_events:
            df_reopen = pd.DataFrame(reopen_events)
            df = pd.concat([df, df_reopen], ignore_index=True)
            df = df.sort_values(by=[class="hl-string">'ServiceRequest', class="hl-string">'EventTime']).reset_index(drop=True)
        
        class=class="hl-string">"hl-comment"># --- Save to CSV ---
        output_filename = class="hl-string">'genesys_event_log.csv'
        df.to_csv(output_filename, index=False, date_format=class="hl-string">'%Y-%m-%dT%H:%M:%S.%fZ')
        print(fclass="hl-string">"Extraction complete. Event log saved to {output_filename}")

    except PureCloudPlatformClientV2.ApiException as e:
        print(fclass="hl-string">"Error calling Genesys Cloud API: {e}")
    except Exception as e:
        print(fclass="hl-string">"An unexpected error occurred: {e}")

if __name__ == class="hl-string">'__main__':
    main()

Steps

Navigate to the Analytics Workspace: Log in to Genesys Cloud CX. Go to the menu and select Admin, then under Analytics, choose Workspace.
Create a New View: Click the plus icon to add a new tab and select "Create a new view". Choose "Interactions" as the view type. This will open the interaction view editor.
Define the Date Range: In the filter panel on the left, set the desired date and time interval for the data you want to extract. It is recommended to start with a smaller date range, such as one week, to test the process.
Select Columns for the Export: This is the most critical step for data preparation. Click on the "Columns" button. From the available fields, select and add the following columns to your view. These fields contain the necessary data for building the event log:
- Conversation ID
- Conversation Start
- Conversation End
- Media Type
- Participants (This is a complex field containing nested agent, queue, and customer data)
- Metrics.tAnswered
- Metrics.tACW
- Metrics.tHeld
- Surveys
Apply Additional Filters: Use the filter panel to narrow down the data if necessary. You can filter by Media Type (e.g., voice, chat), Queue, Direction (inbound/outbound), or other conversation properties to focus on the specific service request process.
Save and Run the View: Save the view with a descriptive name, for example, "Process Mining Service Request Export". The view will then load the data matching your criteria.
Initiate the Export: Once the data is loaded in your view, click the "Export" button, usually represented by a download icon. Select "Export as CSV" or "Export as JSON". CSV is often easier for initial processing.
Download the Exported File: The export job will run in the background. You can monitor its progress by navigating to Performance > My Exports. Once completed, download the file to your local system.
Transform the Raw Data: The exported file has one row per conversation, with participant and segment data nested within a single column. You must run a transformation script, like the Python script provided in the Query section, to un-nest this data. This script will parse the raw export and create a new file where each row represents a single activity, such as "Agent Accepted Interaction" or "Hold Placed on Interaction".
Implement Custom Logic: As part of the transformation, implement any business-specific logic. For example, the "Service Request Re-Opened" activity must be calculated by analyzing subsequent interactions from the same customer within a defined timeframe.
Finalize the Event Log: The transformation script should output a CSV file with columns matching the required format for ProcessMind, including ServiceRequest, ActivityName, EventTime, and other attributes. Ensure all timestamps are in a consistent format.
Upload to ProcessMind: The final, transformed CSV file is now ready to be uploaded to your process mining tool.

Configuration

Permissions: Your Genesys Cloud CX user role must have the Analytics > Conversation > View permission to access the Analytics Workspace and create interaction views.
Date Range: For the initial extraction, use a representative period of 3-6 months. For very high-volume contact centers, it is recommended to export data in smaller chunks, such as weekly or monthly, to avoid hitting system limits and to manage file sizes.
Key Filters:
- Media Type: Filter for the channels relevant to your customer service process, such as voice, chat, or email.
- Queue: If the process is handled by specific teams, filter by the relevant Queue IDs to ensure you are only extracting data for the desired process.
- Direction: Typically, you will want to filter for inbound interactions to capture customer-initiated service requests.
Export Columns: The Participants column is essential as it contains the detailed timeline of events for agents and queues. Without this column, you cannot extract individual activities. Ensure Conversation ID, Conversation Start, and Conversation End are also included.
Performance Considerations: Exporting a large number of columns or a very long date range can cause the export job to take a significant amount of time. If the job fails or times out, reduce the date range and try again. Genesys Cloud has a limit of approximately 1 million conversations per export job.

a Sample Query config

import pandas as pd
import json
from datetime import datetime, timezone

# This script transforms a Genesys Cloud Conversation Detail Export into a process mining event log.
# It assumes the raw data is in a CSV file named 'genesys_export.csv'.

def transform_genesys_export(file_path):
    # Load the raw export file
    try:
        df = pd.read_csv(file_path)
    except FileNotFoundError:
        print(f"Error: The file {file_path} was not found.")
        return None

    event_log = []
    source_system = 'Genesys Cloud CX'
    last_data_update = datetime.now(timezone.utc).isoformat()

    # Rename columns for easier access
    df.rename(columns={
        'Conversation ID': 'conversationId',
        'Conversation Start': 'conversationStart',
        'Conversation End': 'conversationEnd',
        'Media Type': 'mediaType'
    }, inplace=True)

    for index, row in df.iterrows():
        service_request_id = row['conversationId']
        media_type = row['mediaType']
        
        # Activity: Interaction Started
        event_log.append({
            'ServiceRequest': service_request_id,
            'ActivityName': 'Interaction Started',
            'EventTime': row['conversationStart'],
            'EventEndTime': None,
            'AgentId': None, 'AgentName': None, 'QueueName': None, 'WrapUpCode': None,
            'MediaType': media_type, 'SourceSystem': source_system, 'LastDataUpdate': last_data_update
        })

        # The 'Participants' column contains a JSON string that needs to be parsed
        participants = json.loads(row['Participants'])

        for p in participants:
            participant_id = p.get('participantId')
            user_id = p.get('userId')
            queue_name = p.get('queueName')
            purpose = p.get('purpose')

            for s in p.get('sessions', []):
                wrap_up_code = s.get('wrapUpCode')
                for seg in s.get('segments', []):
                    seg_start = seg.get('segmentStart')
                    seg_end = seg.get('segmentEnd')
                    seg_type = seg.get('segmentType')

                    activity_name = None
                    if purpose == 'acd' and seg_type == 'interact':
                        activity_name = 'Conversation Routed to Queue'
                    elif purpose == 'agent' and seg_type == 'alert':
                        activity_name = 'Agent Offered Interaction'
                    elif purpose == 'agent' and seg_type == 'interacting':
                        activity_name = 'Agent Accepted Interaction'
                    elif purpose == 'agent' and seg_type == 'hold':
                        activity_name = 'Hold Placed on Interaction'
                        # Also create the Hold Removed event
                        event_log.append({
                            'ServiceRequest': service_request_id,
                            'ActivityName': 'Hold Removed from Interaction',
                            'EventTime': seg_end,
                            'EventEndTime': None,
                            'AgentId': user_id, 'AgentName': '[Configure based on your system]', 'QueueName': queue_name, 'WrapUpCode': None,
                            'MediaType': media_type, 'SourceSystem': source_system, 'LastDataUpdate': last_data_update
                        })
                    elif purpose == 'agent' and seg_type == 'acw':
                        activity_name = 'After-Call Work Started'
                        # Also create ACW Ended and Wrap-Up Assigned events
                        if wrap_up_code:
                            event_log.append({
                                'ServiceRequest': service_request_id,
                                'ActivityName': 'Wrap-Up Code Assigned',
                                'EventTime': seg_end, # Assume code assigned at end of ACW
                                'EventEndTime': None,
                                'AgentId': user_id, 'AgentName': '[Configure based on your system]', 'QueueName': queue_name, 'WrapUpCode': wrap_up_code,
                                'MediaType': media_type, 'SourceSystem': source_system, 'LastDataUpdate': last_data_update
                            })
                        event_log.append({
                            'ServiceRequest': service_request_id,
                            'ActivityName': 'After-Call Work Ended',
                            'EventTime': seg_end,
                            'EventEndTime': None,
                            'AgentId': user_id, 'AgentName': '[Configure based on your system]', 'QueueName': queue_name, 'WrapUpCode': wrap_up_code,
                            'MediaType': media_type, 'SourceSystem': source_system, 'LastDataUpdate': last_data_update
                        })
                    
                    if seg.get('disconnectType') == 'transfer':
                        activity_name = 'Interaction Transferred'

                    if activity_name:
                        event_log.append({
                            'ServiceRequest': service_request_id,
                            'ActivityName': activity_name,
                            'EventTime': seg_start,
                            'EventEndTime': seg_end,
                            'AgentId': user_id, 'AgentName': '[Configure based on your system]', 'QueueName': queue_name, 'WrapUpCode': None,
                            'MediaType': media_type, 'SourceSystem': source_system, 'LastDataUpdate': last_data_update
                        })
        
        # Activity: Interaction Disconnected
        event_log.append({
            'ServiceRequest': service_request_id,
            'ActivityName': 'Interaction Disconnected',
            'EventTime': row['conversationEnd'],
            'EventEndTime': None,
            'AgentId': None, 'AgentName': None, 'QueueName': None, 'WrapUpCode': None,
            'MediaType': media_type, 'SourceSystem': source_system, 'LastDataUpdate': last_data_update
        })

        # Activity: Customer Satisfaction Survey Sent (from surveys column)
        if pd.notna(row.get('Surveys')):
            surveys = json.loads(row['Surveys'])
            for survey in surveys:
                if survey.get('surveyStatus') == 'Sent':
                    # Survey sent time is not always directly available, using conversation end as an approximation
                    event_log.append({
                        'ServiceRequest': service_request_id,
                        'ActivityName': 'Customer Satisfaction Survey Sent',
                        'EventTime': survey.get('surveySentDate', row['conversationEnd']),
                        'EventEndTime': None,
                        'AgentId': None, 'AgentName': None, 'QueueName': None, 'WrapUpCode': None,
                        'MediaType': media_type, 'SourceSystem': source_system, 'LastDataUpdate': last_data_update
                    })

    # NOTE: 'Service Request Re-Opened' requires custom cross-conversation logic.
    # This logic would need to be implemented here, analyzing the entire dataset
    # to find subsequent interactions from the same customer within a specific timeframe.

    final_df = pd.DataFrame(event_log)
    # Sort by case and event time to create a chronological log
    final_df.sort_values(by=['ServiceRequest', 'EventTime'], inplace=True)
    
    return final_df

if __name__ == '__main__':
    # Specify the path to your downloaded Genesys export CSV file
    input_file = 'genesys_export.csv'
    output_file = 'genesys_event_log.csv'
    
    event_log_df = transform_genesys_export(input_file)
    
    if event_log_df is not None:
        event_log_df.to_csv(output_file, index=False)
        print(f"Transformation complete. Event log saved to {output_file}")

Your Customer Service Data Template

Your Customer Service Data Template

Customer Service Attributes

Customer Service Activities

Extraction Guides

Genesys Cloud Analytics Conversation Detail Export

Steps

Configuration

a Sample Query config

Python Script using the Genesys Cloud Platform API

Steps

Configuration

a Sample Query python

Event Streaming via AWS EventBridge Integration

Steps

Configuration

a Sample Query config

Ready to Get Started?

End Repeat Contacts: Boost Customer Service Today!

Your Customer Service Data Template

Your Customer Service Data Template

Customer Service Attributes

Customer Service Activities

Extraction Guides

Genesys Cloud Analytics Conversation Detail Export

Steps

Configuration

a Sample Query config

Python Script using the Genesys Cloud Platform API

Steps

Configuration

a Sample Query python

Event Streaming via AWS EventBridge Integration

Steps

Configuration

a Sample Query config

We value your privacy

Ready to Get Started?

End Repeat Contacts: Boost Customer Service Today!