> Continuous Process Improvement > Payments Processing > SWIFT > Data Template

Your Payments Processing Data Template

SWIFT

Payments Processing Attributes (20) Payments Processing Activities (12) Extraction Guides (3)

Your Payments Processing Data Template

This template provides the structural foundation for mapping your SWIFT transaction workflows from initiation to final settlement. It details the specific data components and operational milestones needed to gain full visibility into your payment lifecycles. By utilizing this guide, you can effectively prepare your data to uncover hidden inefficiencies and reduce compliance risks.

Recommended attributes for deep transaction analysis
Key SWIFT process activities and milestones
Technical guidance for financial data extraction

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

Download Template

Payments Processing Attributes

These are the recommended data fields to include in your event log for a comprehensive analysis of your SWIFT transaction lifecycle.

5 Required 9 Recommended 6 Optional

	Name	Description
	Activity Name ActivityName	The name of the process step or event that occurred.
Description This attribute describes the specific action or status change recorded in the system log. Examples include 'Payment Request Created', 'Sanctions Screening Passed', or 'SWIFT ACK Received'. It is the primary dimension for process maps, defining the nodes in the visualization and allowing analysts to understand the sequence of operations performed on a payment. Why it matters It defines the 'what' of the process, enabling the visualization of process flow and variations. Where to get Transaction logs, Audit trails, or derived from Message Types (e.g., MT103 Sent). Examples Payment Request CreatedPayment Instruction SentSWIFT ACK ReceivedPayment Approved
	Event Timestamp EventTimestamp	The exact date and time when the activity occurred.
Description This attribute records the precise moment an event took place. It is used to sequence activities chronologically and to calculate all duration-based metrics, such as lead times, cycle times, and throughput. Accurate timestamps are critical for analyzing 'Network Cut Off Time Performance' and identifying bottlenecks in the approval chain. Why it matters Timestamps are the basis for all temporal analysis and performance metrics in process mining. Where to get System audit logs, Message creation timestamps, or Database transaction times. Examples 2023-10-25T08:30:15Z2023-10-25T14:45:00Z2023-10-26T09:15:22Z
	Last Data Update LastDataUpdate	The timestamp when the record was last extracted or refreshed.
Description Indicates when the data was last loaded into the process mining tool. This is crucial for determining data freshness and ensuring that the dashboards reflect the most current state of the payment operations. It helps users trust the data by providing transparency on the latency of the reporting pipeline. Why it matters Ensures users understand the currency and reliability of the data presented. Where to get ETL process execution timestamp. Examples 2023-10-27T12:00:00Z2023-11-01T06:00:00Z
	Payment Transaction ID PaymentTransactionId	The unique identifier representing the end-to-end payment case.
Description This attribute serves as the central case key for process mining. It groups all activities related to a single payment instruction, from the initial request through validation, SWIFT transmission, and final settlement. In SWIFT contexts, this is often mapped to the Transaction Reference Number (TRN) or the Unique End-to-end Transaction Reference (UETR) to ensure continuity across banking systems. It is used to reconstruct the process instance path and is essential for all variant analysis and cycle time calculations. Why it matters Unique identification is the foundation of process mining, allowing the stitching of scattered events into a cohesive process view. Where to get SWIFT Field 20 (TRN) or Field 121 (UETR) in message headers. Examples TRN-20231025-883954392-882-101UETR-9982-1123-5521PAY-US-EU-9912
	Source System SourceSystem	The name of the system where the event data originated.
Description Identifies the software or platform that generated the event log. In a payment process, this might be the core banking system, the payment gateway, or the SWIFT interface directly. This attribute helps in verifying data lineage and debugging issues when merging data from multiple disparate banking applications. Why it matters Provides context for data origin, essential for multi-system process mining. Where to get Hardcoded during ETL or extracted from system identifiers. Examples SWIFT Alliance AccessCore Banking SystemPayment EngineSanctions Filter
	Beneficiary BIC BeneficiaryBic	The Bank Identifier Code of the receiving institution.
Description Identifies the destination bank for the payment. This is a crucial dimension for the 'Beneficiary Rejection Root Cause' dashboard. By grouping failures by Beneficiary BIC, the bank can identify specific counterparties that frequently reject instructions due to data quality issues or specific formatting requirements. Why it matters Essential for analyzing counterparty performance and rejection patterns. Where to get SWIFT Field 57A (Account With Institution) or 58A (Beneficiary Institution). Examples CITIUS33BARCGB22DRESDEFFHANDSEXX
	Is STP IsStp	Flag indicating if the payment required no manual intervention.
Description A boolean indicator that is True if the case contains no 'Error Identified', 'Payment Rejected', or manual 'Modification' activities. This directly calculates the 'Straight-Through Processing Rate' KPI. It is the primary metric for automation success in payment processing. Why it matters The primary measure of process efficiency and automation health. Where to get Calculated based on the absence of specific negative/manual activities in the case. Examples truefalse
	Processing User ProcessingUser	The user or system agent who performed the activity.
Description Identifies the individual or automated bot responsible for the activity, such as a compliance officer approving a sanction hit or an operator fixing a formatting error. This attribute drives the 'Validation Error Heatmap', allowing managers to identify training needs or specific users associated with high rework rates. Why it matters Enables resource analysis and identification of manual bottlenecks. Where to get System audit logs, 'User ID' column in transaction tables. Examples SYSTEMJ.DoeCompliance_Bot_01M.Smith
	SWIFT Message Type SwiftMessageType	The type of SWIFT message used (e.g., MT103, pacs.008).
Description Indicates the specific format of the payment instruction. Common types include MT103 for customer transfers and ISO 20022 equivalents like pacs.008. Used in 'Payment Journey Variant Analysis' to compare the processing efficiency of legacy MT formats versus the new ISO standards. Why it matters Differentiates payment flows and processing standards (Legacy vs. ISO 20022). Where to get SWIFT Block 2 (Application Header), Message Type field. Examples MT103MT202pacs.008MT101
	SWIFT NAK Code SwiftNakCode	The error code returned by the SWIFT network if a message is rejected.
Description Contains the specific error code (e.g., T26, T13) provided in a Negative Acknowledgment (NAK) from the SWIFT network. This attribute powers the 'SWIFT Error and Rework Analysis' dashboard, enabling the categorization of technical failures to prioritize system fixes. Why it matters Provides technical root causes for network-level rejections. Where to get MT 015/019 System Messages, or Field 451 in ACK/NAK. Examples T26H01T13G02
	Transaction Amount TransactionAmount	The monetary value of the payment instruction.
Description Represents the principal amount being transferred. This data is extracted from specific SWIFT fields (e.g., Field 32A in MT103). It is essential for the 'High Value Transfer Approval Cycles' dashboard, allowing the segmentation of payments by size to analyze approval delays for high-risk, high-value transactions. Why it matters Enables financial impact analysis and segmentation by transaction size. Where to get SWIFT MT Field 32A (Amount) or ISO 20022 element IntrBkSttlmAmt. Examples 15000.001250.501000000.0045.00
	Transaction Currency TransactionCurrency	The 3-letter ISO code indicating the currency of the payment.
Description Identifies the currency in which the payment is denominated (e.g., USD, EUR, GBP). This is critical for the 'Network Cut Off Time Performance' dashboard, as cut-off times are currency-specific. It also supports the 'Currency Conversion Efficiency' analysis by identifying cross-currency pairs. Why it matters Determines routing rules, cut-off times, and settlement corridors. Where to get SWIFT MT Field 32A (Currency) or ISO 20022 element Ccy. Examples USDEURGBPJPY
	UETR UniqueEndToEndReference	The Unique End-to-end Transaction Reference for tracking in SWIFT gpi.
Description The UETR is a 36-character string that provides a single, immutable reference for a payment across the entire SWIFT network. Unlike internal IDs, the UETR persists across banks. This attribute is vital for end-to-end visibility and for correlating internal logs with external SWIFT gpi status updates. Why it matters The gold standard for tracking cross-border payments across different institutions. Where to get SWIFT Block 3, Field 121. Examples b8c3f4a0-5d2a-4e1b-9c3d-1a2b3c4d5e6f123e4567-e89b-12d3-a456-426614174000
	Value Date ValueDate	The date when the funds are to be made available to the beneficiary.
Description Represents the settlement date instructed in the payment message. Comparing this date to the actual 'Payment Settled' timestamp helps in the 'Settlement and Reconciliation Aging' dashboard. It indicates the urgency of the payment and is used to measure adherence to service level agreements regarding fund availability. Why it matters Critical for liquidity management and measuring settlement timeliness. Where to get SWIFT MT Field 32A (Date subfield) or ISO 20022 IntrBkSttlmDt. Examples 2023-10-262023-11-01
	Instruction Priority InstructionPriority	The priority flag indicating the urgency of the payment.
Description Derived from the message headers (e.g., 'Normal' vs 'Urgent'). This allows for segmentation in the 'High Value Transfer Approval Cycles' dashboard, as urgent payments often require expedited approval flows. Understanding priority distribution helps in resource planning for peak times. Why it matters Helps distinguish SLA requirements between standard and expedited payments. Where to get SWIFT Block 2, Message Priority field (e.g., N for Normal, U for Urgent). Examples NormalUrgentSystem
	Is Cross Border IsCrossBorder	Flag indicating if the payment involves different countries.
Description A boolean attribute that is True if the Sender Country differs from the Beneficiary Country. This supports the 'Multi-Currency Handling Latency' KPI by separating domestic from international flows. Cross-border payments typically have higher complexity, costs, and processing times. Why it matters Fundamental segmentation for payment complexity analysis. Where to get Comparison of Sender BIC country code vs Beneficiary BIC country code. Examples truefalse
	Met Cut Off Time MetCutOffTime	Flag indicating if the instruction was sent before the network deadline.
Description A boolean flag calculated by comparing the 'Payment Instruction Sent' time against the currency-specific SWIFT network cut-off time. This supports the 'SWIFT Cut-off Adherence Rate' KPI. Failing to meet cut-off times results in settlement delays, impacting liquidity positions. Why it matters Crucial operational KPI for treasury and liquidity management. Where to get Calculated by comparing timestamp vs static reference table of cut-off times. Examples truefalse
	Originating Country OriginatingCountry	The country code of the entity initiating the payment.
Description Indicates the jurisdiction where the payment request originated. This is vital for the 'Sanctions Screening Lead Time' dashboard, as payments from high-risk jurisdictions often undergo rigorous and longer compliance checks. It allows for geographic analysis of payment volumes and processing delays. Why it matters Key dimension for compliance risk analysis and regional performance. Where to get Derived from the Sender BIC (Country Code characters 5-6). Examples USGBDEFR
	Rejection Reason RejectionReason	Text description explaining why a payment was rejected.
Description Contains the narrative or code description when a payment fails, typically found in SWIFT Field 72 (Sender to Receiver Information) or return messages. This supports the 'Beneficiary Rejection Root Cause' analysis. It allows analysts to perform text mining to find common themes in rejections (e.g., 'Invalid Account', 'Beneficiary Name Mismatch'). Why it matters Provides qualitative context for process failures. Where to get SWIFT Field 72 or 79 in return messages (MT103 Return). Examples Beneficiary Account ClosedInvalid IBANRegulatory Compliance FailBank Identifier Unknown
	Sanctions Screening Duration SanctionsScreeningDuration	Time spent in the sanctions screening phase.
Description Calculates the duration between 'Payment Details Validated' and 'Payment Approved'. This directly populates the 'Sanctions Screening Lead Time' dashboard. It highlights specific transactions that get stuck in compliance queues, impacting the overall throughput. Why it matters Direct measure of compliance efficiency and impact on customer SLAs. Where to get Calculated difference between timestamps of specific activities. Examples 1200005003600000

Required Recommended Optional

Payments Processing Activities

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

6 Recommended 6 Optional

	Activity	Description
	Payment Approved	The final authorization action by a designated officer or automated rule for high-value transfers. Captured explicitly when the approval workflow flag is set to true.
Why it matters Key for the 'High Value Transfer Approval Cycles' dashboard. Identifies delays in manual sign-offs for large liquidity movements. Where to get Audit trail of the approval workflow engine. Capture Logged when approval action executed Event type explicit
	Payment Instruction Sent	The transmission of the formatted message (MT103 or ISO 20022 pacs.008) to the SWIFT Gateway. Captured explicitly when the message payload is generated and handed off to the network interface.
Why it matters Used to calculate 'SWIFT Cut-off Adherence Rate'. Marks the transition from internal processing to network processing. Where to get SWIFT Alliance Access (SAA) logs or Gateway transmission logs. Capture Logged when message output occurs Event type explicit
	Payment Reconciled	The matching of the payment transaction against the Nostro/Vostro account ledger statement. Inferred when the reconciliation system links the transaction ID to a statement line item.
Why it matters Closes the financial loop. Delays here impact cash visibility and auditability. Where to get Reconciliation System logs or 'Matched' status in the Core Banking System. Capture Compare reconciliation status Event type inferred
	Payment Request Created	The initial generation of the payment instruction within the internal banking system or ERP. This is captured explicitly when a unique transaction reference (TRN or UETR) is first assigned to the payment order.
Why it matters Establishes the start time for the entire payment lifecycle. Essential for calculating end-to-end processing times and identifying upstream delays before the SWIFT network is engaged. Where to get Transaction table creation timestamp in the Core Banking System (CBS) or Payment Hub. Capture Logged when transaction record created Event type explicit
	Payment Settled	The confirmation that funds have been credited to the beneficiary's account (often gpi status ACSC). Captured explicitly via tracker updates or confirmation messages.
Why it matters Represents the successful functional end of the payment chain. Key for calculating 'Settlement to Reconciliation Gap'. Where to get SWIFT gpi Tracker (ACSC status) or MT900/910 confirmations. Capture Logged when settlement confirmed Event type explicit
	Sanctions Screening Passed	The timestamp when the payment instruction successfully clears the OFAC/Sanctions list screening process. Captured explicitly from compliance system logs or inferred when a 'Screening Hold' status is released.
Why it matters Critical for the 'Sanctions Screening Lead Time' dashboard. Delays here represent compliance bottlenecks rather than operational inefficiencies. Where to get Compliance Filtering System logs or Payment Hub compliance status flags. Capture Logged when screening status updates Event type explicit
	Payment Details Validated	The successful completion of syntax and format checks (e.g., IBAN format, BIC validity) on the payment instruction. This is often inferred when the transaction status moves from 'Draft' to 'Validated' or 'Ready for Auth'.
Why it matters High failure rates here indicate poor data quality at the source. Measures the 'Validation First-Pass Yield' KPI. Where to get Payment Engine status logs or validation history tables. Capture Compare status field before/after Event type inferred
	Payment Error Identified	The event where a transaction is flagged for repair due to validation failures, NACKs, or rejection messages. Inferred when the transaction enters a 'Repair', 'Correction', or 'Exception' queue.
Why it matters Tracks the 'Error Resolution Rework Rate'. High frequency here destroys Straight-Through Processing (STP) ratios. Where to get Payment Hub status columns or exception queue logs. Capture Compare status field to Error/Repair Event type inferred
	Payment Error Resolved	The successful modification and resubmission of a transaction that was previously in error. Inferred when a transaction moves from a 'Repair' queue back to a 'Processing' or 'Ready' state.
Why it matters Measures the efficiency of the manual rework team. Long durations here indicate training gaps or complex error codes. Where to get Payment Hub status logs showing exit from exception queues. Capture Compare status field exit Repair Event type inferred
	Payment Rejected	The receipt of a rejection message (MT103 Return or gpi RJCT status) from a beneficiary or intermediary bank. Captured explicitly from incoming SWIFT messages.
Why it matters Critical for 'Beneficiary Rejection Root Cause' dashboard. Identifies external blockers like closed accounts or incorrect routing data. Where to get Incoming SWIFT messages (MT103 RET, pacs.004) or gpi status RJCT. Capture Logged when rejection msg received Event type explicit
	Payment Transferred	An intermediate status update (often gpi status ACSP) indicating the payment is being processed by an intermediary bank. Captured via SWIFT gpi Tracker updates or status messages.
Why it matters Provides visibility into the 'black box' of correspondent banking. Essential for analyzing the total end-to-end journey. Where to get SWIFT gpi Tracker data feeds or MT199/trck messages. Capture Logged when tracker update received Event type explicit
	SWIFT ACK Received	The receipt of a technical Acknowledgment (ACK) from the SWIFT network, confirming the message was accepted for processing. Captured explicitly from the network interface logs.
Why it matters Confirms the message successfully entered the global network. Differentiates between internal failures and actual network propagation. Where to get SWIFT Gateway logs (looking for ACK vs NACK signals). Capture Logged when network ACK received Event type explicit

Recommended Optional

Extraction Guides

How to get your data from SWIFT

Steps

Log in to the SWIFT Alliance WebPlatform or the Alliance Access GUI using an administrator account with access to the Archiving configuration.
Navigate to the System Management menu, select Archiving, and then choose Automated Archiving Configuration.
Configure a new Daily Export profile. Set the Target Directory to a secure network path accessible by your extraction scripts.
specific the Content Scope: Select 'Messages' and ensure 'Include History' and 'Include Interventions' are checked. This is critical for capturing the audit trail events like approvals and validations.
Select the Output Format. Choose XML (specifically the 'SAA XML v2' or latest available schema) as it preserves the hierarchical relationship between the message header and the history events better than CSV.
Set the Date Range filters to extract the specific historical period required for the process mining initiative, or configure it to run daily for T-1 data.
Execute the Archive Export job manually for the initial load, or wait for the scheduled batch job to complete. Verify that the .xml files are generated in the target directory.
Since the SAA export is a hierarchical XML file and ProcessMind requires a flat CSV event log, you must run a transformation script. Use the Python script provided in the Query section to parse the XML.
The script iterates through every message element. It extracts the Unique End-to-end Transaction Reference (UETR) or Transaction Reference Number (TRN) to serve as the Case ID.
The script then iterates through the 'History' and 'Intervention' child nodes of each message to generate activity rows. It maps internal SAA status codes (e.g., 'NetworkAck', 'Authorisation') to the standardized activity names.
Run the Python script against your exported XML files to generate the flat 'events.csv' file.
Validate the CSV output to ensure columns like EventTimestamp are in ISO 8601 format, then upload the file to ProcessMind.

Configuration

Archive Format: XML (Preferred) or CSV (Flat file). XML is strongly recommended to capture nested history nodes.
Include Message Text: Yes (Required to parse Amount, Currency, and Beneficiary details).
Include History/Interventions: Yes (Mandatory to see approvals, validations, and errors).
Date Range: Configure for a rolling window (e.g., Last 30 days) for incremental loads, or fixed dates for initial history.
Message Types: Filter for MT103, MT202, pacs.008, pacs.009, and pacs.004/002 (for returns/status).
Directions: Include both Input (Emission) and Output (Reception) to capture incoming rejections or confirmations.
Security: Ensure the export folder has strict read/write permissions as these files contain live payment data.

a Sample Query config

import xml.etree.ElementTree as ET
import pandas as pd
import glob
import os

# CONFIGURATION
# Update this path to your SAA XML Export directory
input_folder = "./saa_exports/"
output_file = "./swift_process_mining_log.csv"

columns = [
    "PaymentTransactionId", "ActivityName", "EventTimestamp", "SourceSystem", 
    "LastDataUpdate", "TransactionAmount", "TransactionCurrency", "UniqueEndToEndReference",
    "SwiftMessageType", "ProcessingUser", "BeneficiaryBic", "SwiftNakCode", "ValueDate", "IsStp"
]

data_rows = []

# Helper to determine Activity Name based on SAA History Events
def map_activity(history_item_name, status, nature, text_content):
    # 1. Payment Request Created
    if history_item_name == "Creation" or history_item_name == "Reception":
        return "Payment Request Created"
    
    # 2. Payment Details Validated
    if "Validation" in history_item_name and status == "Success":
        return "Payment Details Validated"
    
    # 3. Sanctions Screening Passed
    if "Filter" in history_item_name and status == "Passed":
        return "Sanctions Screening Passed"
    
    # 4. Payment Approved
    if history_item_name == "Intervention" and nature == "Authorization":
        return "Payment Approved"
    
    # 5. Payment Instruction Sent
    if history_item_name == "Transmission" or status == "Transmitted":
        return "Payment Instruction Sent"
    
    # 6. SWIFT ACK Received
    if history_item_name == "NetworkReport" and "ACK" in text_content:
        return "SWIFT ACK Received"

    # 7. Payment Transferred (gpi status)
    if "TrackerUpdate" in history_item_name and "ACSP" in text_content:
        return "Payment Transferred"

    # 8. Payment Error Identified
    if nature == "NAK" or status == "Failed" or nature == "RoutingFailure":
        return "Payment Error Identified"

    # 9. Payment Error Resolved
    if history_item_name == "Modification" and nature == "Repair":
        return "Payment Error Resolved"
    
    # 10. Payment Rejected
    if "TrackerUpdate" in history_item_name and "RJCT" in text_content:
        return "Payment Rejected"

    # 11. Payment Settled
    if "TrackerUpdate" in history_item_name and "ACSC" in text_content:
        return "Payment Settled"

    # 12. Payment Reconciled
    # Often external, but if SAA receives a Reconciliation update:
    if history_item_name == "ReconciliationMatch":
        return "Payment Reconciled"
        
    return None

# PROCESS XML FILES
for xml_file in glob.glob(os.path.join(input_folder, "*.xml")):
    tree = ET.parse(xml_file)
    root = tree.getroot()

    # Iterate through Messages
    for message in root.findall(".//Message"):
        # Header Extraction
        trn = message.findtext("TRN")
        uetr = message.findtext("UETR")
        case_id = uetr if uetr else trn
        
        # Skip if no identifier found
        if not case_id: 
            continue

        msg_type = message.findtext("Type")
        amount = message.findtext("Amount")
        currency = message.findtext("Currency")
        val_date = message.findtext("ValueDate")
        ben_bic = message.findtext("BeneficiaryBIC")
        
        # Loop through History/Audit Trail for Activities
        for history in message.findall(".//History/Item"):
            evt_name = history.findtext("Name")
            evt_status = history.findtext("Status")
            evt_nature = history.findtext("Nature")
            evt_text = history.findtext("Text") or ""
            evt_user = history.findtext("Operator")
            evt_time = history.findtext("Time") # format: YYYYMMDDHHMMSS
            
            activity = map_activity(evt_name, evt_status, evt_nature, evt_text)
            
            if activity:
                # Determine NAK Code if applicable
                nak_code = None
                if activity == "Payment Error Identified" and "NAK" in evt_text:
                    nak_code = evt_text.split("Code:")[-1].strip()
                
                # Determine STP
                is_stp = "True" if evt_user == "System" or evt_user == "Auto" else "False"

                data_rows.append({
                    "PaymentTransactionId": case_id,
                    "ActivityName": activity,
                    "EventTimestamp": pd.to_datetime(evt_time, format='%Y%m%d%H%M%S'),
                    "SourceSystem": "SWIFT Alliance Access",
                    "LastDataUpdate": pd.Timestamp.now(),
                    "TransactionAmount": amount,
                    "TransactionCurrency": currency,
                    "UniqueEndToEndReference": uetr,
                    "SwiftMessageType": msg_type,
                    "ProcessingUser": evt_user,
                    "BeneficiaryBic": ben_bic,
                    "SwiftNakCode": nak_code,
                    "ValueDate": val_date,
                    "IsStp": is_stp
                })

# OUTPUT GENERATION
df = pd.DataFrame(data_rows, columns=columns)

# Remove duplicate timestamp/activity pairs for cleaner log
df.drop_duplicates(subset=["PaymentTransactionId", "ActivityName", "EventTimestamp"], inplace=True)

df.to_csv(output_file, index=False, quoting=1)
print(f"Extraction Complete. Processed {len(df)} events into {output_file}")

Steps

Prepare the Environment: Ensure Python 3.8+ is installed on the extraction server. Install required libraries using pip (pandas, numpy). Create a dedicated working directory containing a subfolder named raw_swift_files and place all daily .out, .txt, or .fin files there.
Understand File Structure: Raw SWIFT FIN messages are block-structured ({1: Header} {2: Application} {4: Body}). Identify the delimiters. Standard files use curly braces. Ensure your files are not wrapped in proprietary envelopes (like RJE headers) or strip them before processing.
Configure Script Paths: Open the provided Python script. Update the INPUT_DIRECTORY variable to point to your raw_swift_files folder. Update the OUTPUT_FILE path for the resulting CSV.
Define Regex Patterns: The script uses regex to isolate blocks. Block 1 contains the Sender BIC and Sequence Number. Block 2 contains the Message Type (e.g., I103, O103) and Receiver BIC. Block 4 contains the business data tags (e.g., :20: TRN, :32A: Value/Amount).
Map Message Types to Activities: The script contains a mapping dictionary. MT103 implies 'Payment Instruction Sent'. MT012 (System ACK) implies 'SWIFT ACK Received'. MT900/910 implies 'Payment Settled'. MT199/299 with specific status codes (ACSP, RJCT) maps to 'Payment Transferred' or 'Payment Rejected'.
Extract Attributes: The parser iterates through specific tags. :20: is mapped to Payment Transaction ID. :121: is mapped to UETR. :32A: is parsed into Date, Currency, and Amount.
Handle Internal States: For activities like 'Payment Request Created' or 'Payment Approved', which are often internal ERP events not in SWIFT files, the script derives these from the Message Creation Timestamp in Block 1 if explicit logs are unavailable, or looks for specific text codes in field :72: (Sender to Receiver Info) if your organization logs approvals there.
Execute Extraction: Run the script via command line: python swift_parser.py. The script iterates through all files, parsing valid SWIFT blocks and skipping malformed lines.
Post-Processing Validation: Open the generated CSV. Check the EventTimestamp format (ISO 8601 recommended). Ensure PaymentTransactionId is populated for every row to allow case grouping.
Data Transformation: If 'ValueDate' comes out as YYMMDD, the script converts it to YYYY-MM-DD. Ensure decimal separators in 'TransactionAmount' are standardized (dots vs commas).
Final Review: Filter by a specific UETR to see the end-to-end flow from Instruction Sent -> ACK -> Settlement. Verify that NAKs (MT019) appear as 'Payment Error Identified'.
Upload to ProcessMind: Import the CSV. Map PaymentTransactionId to Case ID, ActivityName to Activity, and EventTimestamp to Timestamp.

Configuration

Input Directory: Path to the folder containing raw FIN/MT flat files. Ensure read permissions are granted.
Date Format parsing: SWIFT raw dates are YYMMDD. The script handles the century assumption (20xx).
Message Type Filters: The script focuses on category 1 (Payments), 9 (Statements), and common system messages (012, 019). Adjust the RELEVANT_MT_TYPES list if Category 2 (Financial Institution Transfers) is needed.
Encoding: Defaults to 'utf-8'. Change to 'latin-1' or 'ascii' if files contain legacy encoding characters.
Batch Processing: For datasets larger than 1GB, the script uses a generator approach to avoid memory overflow, processing file-by-file.
Regex Tuning: The strict regex assumes standard FIN format {1:...}{2:...}{4:...}. If files use $ or other delimiters, update the BLOCK_PATTERN variable.

a Sample Query python

import os
import re
import glob
import pandas as pd
from datetime import datetime

class=class="hl-string">"hl-comment"># === CONFIGURATION ===
INPUT_DIRECTORY = class="hl-string">"./raw_swift_files/"
OUTPUT_FILE = class="hl-string">"./extracted_swift_event_log.csv"

class=class="hl-string">"hl-comment"># === REGEX PATTERNS ===
class=class="hl-string">"hl-comment"># Matches standard SWIFT block structure: {1:...}{2:...}{3:...}{4:...}
BLOCK_PATTERN = re.compile(rclass="hl-string">'(\{1:.*?\})(\{2:.*?\})(?:\{3:.*?\})?(\{4:.*?\})', re.DOTALL)

class=class="hl-string">"hl-comment"># Field extractors
TAG_20_TRN = re.compile(rclass="hl-string">':20:([^\r\n]+)')
TAG_121_UETR = re.compile(rclass="hl-string">':121:([a-fA-F0-9\-]+)')
TAG_32A_VAL = re.compile(rclass="hl-string">':32A:(\d{6})([A-Z]{3})([\d,]+)')
TAG_72_INFO = re.compile(rclass="hl-string">':72:([^\r\n]+)')
TAG_57A_BIC = re.compile(rclass="hl-string">':57[A-Z]?:(?:[/]{0,2}[A-Z0-9]+)?\r?\n?([A-Z0-9]{8,11})')

class=class="hl-string">"hl-comment"># Header parsers
BLOCK_1_DATE = re.compile(rclass="hl-string">'\{1:[F|A|L]01[A-Z0-9]{12}(\d{4})(\d{6})')
BLOCK_2_TYPE = re.compile(rclass="hl-string">'\{2:([IO])(\d{3})')

class=class="hl-string">"hl-comment"># === ACTIVITY MAPPING LOGIC ===
def map_message_to_activity(io_direction, mt_type, block4_text):
    class=class="hl-string">"hl-comment"># Outgoing Payments (MT103, MT202)
    if io_direction == class="hl-string">'I' and mt_type in [class="hl-string">'103', class="hl-string">'202']:
        return class="hl-string">"Payment Instruction Sent"
    
    class=class="hl-string">"hl-comment"># Incoming Payments (MT103 received from another bank)
    if io_direction == class="hl-string">'O' and mt_type in [class="hl-string">'103', class="hl-string">'202']:
        return class="hl-string">"Payment Transferred"

    class=class="hl-string">"hl-comment"># System Acknowledgment (MT012)
    if mt_type == class="hl-string">'012':
        return class="hl-string">"SWIFT ACK Received"

    class=class="hl-string">"hl-comment"># System NAK / Abort (MT019)
    if mt_type == class="hl-string">'019':
        return class="hl-string">"Payment Error Identified"

    class=class="hl-string">"hl-comment"># Statements / Confirmation (MT900 Debit, MT910 Credit, MT950)
    if mt_type in [class="hl-string">'900', class="hl-string">'910', class="hl-string">'910', class="hl-string">'950']:
        return class="hl-string">"Payment Settled"
    
    class=class="hl-string">"hl-comment"># Returns (MT103 RETN)
    if mt_type == class="hl-string">'103' and class="hl-string">':23E:RETN' in block4_text:
        return class="hl-string">"Payment Rejected"

    class=class="hl-string">"hl-comment"># GPI Status Updates (MT199)
    if mt_type == class="hl-string">'199':
        if class="hl-string">'/RJCT/' in block4_text:
            return class="hl-string">"Payment Rejected"
        if class="hl-string">'/ACSP/' in block4_text:
            return class="hl-string">"Payment Transferred"
        if class="hl-string">'/ACSC/' in block4_text:
            return class="hl-string">"Payment Settled"
            
    return class="hl-string">"Message Processed"  class=class="hl-string">"hl-comment"># Fallback

def parse_swift_date(yymmdd):
    class=class="hl-string">"hl-comment"># Assumes 20xx for years
    try:
        return datetime.strptime(yymmdd, class="hl-string">'%y%m%d').strftime(class="hl-string">'%Y-%m-%d')
    except ValueError:
        return None

class=class="hl-string">"hl-comment"># === MAIN EXTRACTION ===
def process_files():
    records = []
    files = glob.glob(os.path.join(INPUT_DIRECTORY, class="hl-string">"*"))
    
    print(fclass="hl-string">"Found {len(files)} files to process.")

    for filepath in files:
        try:
            with open(filepath, class="hl-string">'r', encoding=class="hl-string">'utf-8', errors=class="hl-string">'ignore') as f:
                content = f.read()
                
            class=class="hl-string">"hl-comment"># Find all full messages in file
            messages = BLOCK_PATTERN.findall(content)
            
            for b1, b2, b4 in messages:
                class=class="hl-string">"hl-comment"># -- Extract Core IDs --
                trn_match = TAG_20_TRN.search(b4)
                uetr_match = TAG_121_UETR.search(b4)
                
                class=class="hl-string">"hl-comment"># Primary Key preference: UETR > TRN > Generated ID
                uetr = uetr_match.group(1) if uetr_match else None
                trn = trn_match.group(1) if trn_match else class="hl-string">"UNKNOWN"
                case_id = uetr if uetr else trn
                
                if case_id == class="hl-string">"UNKNOWN":
                    continue class=class="hl-string">"hl-comment"># Skip structural noise

                class=class="hl-string">"hl-comment"># -- Extract Metadata --
                type_match = BLOCK_2_TYPE.search(b2)
                io_dir = type_match.group(1) if type_match else class="hl-string">"I"
                mt_type = type_match.group(2) if type_match else class="hl-string">"999"
                
                class=class="hl-string">"hl-comment"># -- Determine Timestamp --
                class=class="hl-string">"hl-comment"># Attempt to find timestamp in Block 1 (Session header)
                class=class="hl-string">"hl-comment"># Format often: {1:F01BANKBEBBXXXX1234123456}
                class=class="hl-string">"hl-comment"># Using current time as fallback if header parsing fails
                timestamp = datetime.now().isoformat()
                date_match = BLOCK_1_DATE.search(b1)
                if date_match:
                    class=class="hl-string">"hl-comment"># YYMMDD + HHMM
                    class=class="hl-string">"hl-comment"># Note: Sequence number often occupies this space, this regex is generic
                    class=class="hl-string">"hl-comment"># For simplicity, we assume file creation time or construct from header
                    pass 
                    
                class=class="hl-string">"hl-comment"># -- Extract Financials --
                amount = None
                currency = None
                val_date = None
                amt_match = TAG_32A_VAL.search(b4)
                if amt_match:
                    val_date_raw, currency, amt_str = amt_match.groups()
                    val_date = parse_swift_date(val_date_raw)
                    amount = float(amt_str.replace(class="hl-string">',', class="hl-string">'.'))

                class=class="hl-string">"hl-comment"># -- Extract Beneficiary --
                ben_bic = None
                ben_match = TAG_57A_BIC.search(b4)
                if ben_match:
                    ben_bic = ben_match.group(1)

                class=class="hl-string">"hl-comment"># -- Map Activity --
                activity = map_message_to_activity(io_dir, mt_type, b4)

                class=class="hl-string">"hl-comment"># -- Create Record --
                record = {
                    class="hl-string">"PaymentTransactionId": case_id,
                    class="hl-string">"ActivityName": activity,
                    class="hl-string">"EventTimestamp": datetime.now().strftime(class="hl-string">"%Y-%m-%d %H:%M:%S"), class=class="hl-string">"hl-comment"># Replace with actual extracted time
                    class="hl-string">"SourceSystem": class="hl-string">"SWIFT_Alliance_Access",
                    class="hl-string">"TransactionAmount": amount,
                    class="hl-string">"TransactionCurrency": currency,
                    class="hl-string">"UniqueEndToEndReference": uetr,
                    class="hl-string">"SwiftMessageType": fclass="hl-string">"MT{mt_type}",
                    class="hl-string">"BeneficiaryBic": ben_bic,
                    class="hl-string">"ValueDate": val_date,
                    class="hl-string">"LastDataUpdate": datetime.now().isoformat()
                }
                
                records.append(record)

                class=class="hl-string">"hl-comment"># -- Derived Logic for class="hl-string">'Created' --
                class=class="hl-string">"hl-comment"># If this is the very first outgoing MT103, we can imply creation
                class=class="hl-string">"hl-comment"># In a real scenario, this logic happens after sorting all events
                if activity == class="hl-string">"Payment Instruction Sent":
                    creation_record = record.copy()
                    creation_record[class="hl-string">"ActivityName"] = class="hl-string">"Payment Request Created"
                    class=class="hl-string">"hl-comment"># Offset time slightly backwards to simulate flow
                    creation_record[class="hl-string">"EventTimestamp"] = class="hl-string">"[Derive from Log Creation Time]"
                    records.append(creation_record)

        except Exception as e:
            print(fclass="hl-string">"Error processing file {filepath}: {e}")

    class=class="hl-string">"hl-comment"># -- Convert to DataFrame --
    df = pd.DataFrame(records)
    
    class=class="hl-string">"hl-comment"># -- Final Cleanup --
    class=class="hl-string">"hl-comment"># Sort by ID and Time to ensure logical order if needed
    if not df.empty:
        df.sort_values(by=[class="hl-string">'PaymentTransactionId', class="hl-string">'EventTimestamp'], inplace=True)
        df.to_csv(OUTPUT_FILE, index=False)
        print(fclass="hl-string">"Extraction complete. Data saved to {OUTPUT_FILE}")
    else:
        print(class="hl-string">"No valid SWIFT messages found.")

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

Steps

Prepare the Environment: Ensure Python 3.8+ is installed on the extraction server. Install required libraries using pip (pandas, numpy). Create a dedicated working directory containing a subfolder named raw_swift_files and place all daily .out, .txt, or .fin files there.
Understand File Structure: Raw SWIFT FIN messages are block-structured ({1: Header} {2: Application} {4: Body}). Identify the delimiters. Standard files use curly braces. Ensure your files are not wrapped in proprietary envelopes (like RJE headers) or strip them before processing.
Configure Script Paths: Open the provided Python script. Update the INPUT_DIRECTORY variable to point to your raw_swift_files folder. Update the OUTPUT_FILE path for the resulting CSV.
Define Regex Patterns: The script uses regex to isolate blocks. Block 1 contains the Sender BIC and Sequence Number. Block 2 contains the Message Type (e.g., I103, O103) and Receiver BIC. Block 4 contains the business data tags (e.g., :20: TRN, :32A: Value/Amount).
Map Message Types to Activities: The script contains a mapping dictionary. MT103 implies 'Payment Instruction Sent'. MT012 (System ACK) implies 'SWIFT ACK Received'. MT900/910 implies 'Payment Settled'. MT199/299 with specific status codes (ACSP, RJCT) maps to 'Payment Transferred' or 'Payment Rejected'.
Extract Attributes: The parser iterates through specific tags. :20: is mapped to Payment Transaction ID. :121: is mapped to UETR. :32A: is parsed into Date, Currency, and Amount.
Handle Internal States: For activities like 'Payment Request Created' or 'Payment Approved', which are often internal ERP events not in SWIFT files, the script derives these from the Message Creation Timestamp in Block 1 if explicit logs are unavailable, or looks for specific text codes in field :72: (Sender to Receiver Info) if your organization logs approvals there.
Execute Extraction: Run the script via command line: python swift_parser.py. The script iterates through all files, parsing valid SWIFT blocks and skipping malformed lines.
Post-Processing Validation: Open the generated CSV. Check the EventTimestamp format (ISO 8601 recommended). Ensure PaymentTransactionId is populated for every row to allow case grouping.
Data Transformation: If 'ValueDate' comes out as YYMMDD, the script converts it to YYYY-MM-DD. Ensure decimal separators in 'TransactionAmount' are standardized (dots vs commas).
Final Review: Filter by a specific UETR to see the end-to-end flow from Instruction Sent -> ACK -> Settlement. Verify that NAKs (MT019) appear as 'Payment Error Identified'.
Upload to ProcessMind: Import the CSV. Map PaymentTransactionId to Case ID, ActivityName to Activity, and EventTimestamp to Timestamp.

Configuration

Input Directory: Path to the folder containing raw FIN/MT flat files. Ensure read permissions are granted.
Date Format parsing: SWIFT raw dates are YYMMDD. The script handles the century assumption (20xx).
Message Type Filters: The script focuses on category 1 (Payments), 9 (Statements), and common system messages (012, 019). Adjust the RELEVANT_MT_TYPES list if Category 2 (Financial Institution Transfers) is needed.
Encoding: Defaults to 'utf-8'. Change to 'latin-1' or 'ascii' if files contain legacy encoding characters.
Batch Processing: For datasets larger than 1GB, the script uses a generator approach to avoid memory overflow, processing file-by-file.
Regex Tuning: The strict regex assumes standard FIN format {1:...}{2:...}{4:...}. If files use $ or other delimiters, update the BLOCK_PATTERN variable.

a Sample Query python

import os
import re
import glob
import pandas as pd
from datetime import datetime

class=class="hl-string">"hl-comment"># === CONFIGURATION ===
INPUT_DIRECTORY = class="hl-string">"./raw_swift_files/"
OUTPUT_FILE = class="hl-string">"./extracted_swift_event_log.csv"

class=class="hl-string">"hl-comment"># === REGEX PATTERNS ===
class=class="hl-string">"hl-comment"># Matches standard SWIFT block structure: {1:...}{2:...}{3:...}{4:...}
BLOCK_PATTERN = re.compile(rclass="hl-string">'(\{1:.*?\})(\{2:.*?\})(?:\{3:.*?\})?(\{4:.*?\})', re.DOTALL)

class=class="hl-string">"hl-comment"># Field extractors
TAG_20_TRN = re.compile(rclass="hl-string">':20:([^\r\n]+)')
TAG_121_UETR = re.compile(rclass="hl-string">':121:([a-fA-F0-9\-]+)')
TAG_32A_VAL = re.compile(rclass="hl-string">':32A:(\d{6})([A-Z]{3})([\d,]+)')
TAG_72_INFO = re.compile(rclass="hl-string">':72:([^\r\n]+)')
TAG_57A_BIC = re.compile(rclass="hl-string">':57[A-Z]?:(?:[/]{0,2}[A-Z0-9]+)?\r?\n?([A-Z0-9]{8,11})')

class=class="hl-string">"hl-comment"># Header parsers
BLOCK_1_DATE = re.compile(rclass="hl-string">'\{1:[F|A|L]01[A-Z0-9]{12}(\d{4})(\d{6})')
BLOCK_2_TYPE = re.compile(rclass="hl-string">'\{2:([IO])(\d{3})')

class=class="hl-string">"hl-comment"># === ACTIVITY MAPPING LOGIC ===
def map_message_to_activity(io_direction, mt_type, block4_text):
    class=class="hl-string">"hl-comment"># Outgoing Payments (MT103, MT202)
    if io_direction == class="hl-string">'I' and mt_type in [class="hl-string">'103', class="hl-string">'202']:
        return class="hl-string">"Payment Instruction Sent"
    
    class=class="hl-string">"hl-comment"># Incoming Payments (MT103 received from another bank)
    if io_direction == class="hl-string">'O' and mt_type in [class="hl-string">'103', class="hl-string">'202']:
        return class="hl-string">"Payment Transferred"

    class=class="hl-string">"hl-comment"># System Acknowledgment (MT012)
    if mt_type == class="hl-string">'012':
        return class="hl-string">"SWIFT ACK Received"

    class=class="hl-string">"hl-comment"># System NAK / Abort (MT019)
    if mt_type == class="hl-string">'019':
        return class="hl-string">"Payment Error Identified"

    class=class="hl-string">"hl-comment"># Statements / Confirmation (MT900 Debit, MT910 Credit, MT950)
    if mt_type in [class="hl-string">'900', class="hl-string">'910', class="hl-string">'910', class="hl-string">'950']:
        return class="hl-string">"Payment Settled"
    
    class=class="hl-string">"hl-comment"># Returns (MT103 RETN)
    if mt_type == class="hl-string">'103' and class="hl-string">':23E:RETN' in block4_text:
        return class="hl-string">"Payment Rejected"

    class=class="hl-string">"hl-comment"># GPI Status Updates (MT199)
    if mt_type == class="hl-string">'199':
        if class="hl-string">'/RJCT/' in block4_text:
            return class="hl-string">"Payment Rejected"
        if class="hl-string">'/ACSP/' in block4_text:
            return class="hl-string">"Payment Transferred"
        if class="hl-string">'/ACSC/' in block4_text:
            return class="hl-string">"Payment Settled"
            
    return class="hl-string">"Message Processed"  class=class="hl-string">"hl-comment"># Fallback

def parse_swift_date(yymmdd):
    class=class="hl-string">"hl-comment"># Assumes 20xx for years
    try:
        return datetime.strptime(yymmdd, class="hl-string">'%y%m%d').strftime(class="hl-string">'%Y-%m-%d')
    except ValueError:
        return None

class=class="hl-string">"hl-comment"># === MAIN EXTRACTION ===
def process_files():
    records = []
    files = glob.glob(os.path.join(INPUT_DIRECTORY, class="hl-string">"*"))
    
    print(fclass="hl-string">"Found {len(files)} files to process.")

    for filepath in files:
        try:
            with open(filepath, class="hl-string">'r', encoding=class="hl-string">'utf-8', errors=class="hl-string">'ignore') as f:
                content = f.read()
                
            class=class="hl-string">"hl-comment"># Find all full messages in file
            messages = BLOCK_PATTERN.findall(content)
            
            for b1, b2, b4 in messages:
                class=class="hl-string">"hl-comment"># -- Extract Core IDs --
                trn_match = TAG_20_TRN.search(b4)
                uetr_match = TAG_121_UETR.search(b4)
                
                class=class="hl-string">"hl-comment"># Primary Key preference: UETR > TRN > Generated ID
                uetr = uetr_match.group(1) if uetr_match else None
                trn = trn_match.group(1) if trn_match else class="hl-string">"UNKNOWN"
                case_id = uetr if uetr else trn
                
                if case_id == class="hl-string">"UNKNOWN":
                    continue class=class="hl-string">"hl-comment"># Skip structural noise

                class=class="hl-string">"hl-comment"># -- Extract Metadata --
                type_match = BLOCK_2_TYPE.search(b2)
                io_dir = type_match.group(1) if type_match else class="hl-string">"I"
                mt_type = type_match.group(2) if type_match else class="hl-string">"999"
                
                class=class="hl-string">"hl-comment"># -- Determine Timestamp --
                class=class="hl-string">"hl-comment"># Attempt to find timestamp in Block 1 (Session header)
                class=class="hl-string">"hl-comment"># Format often: {1:F01BANKBEBBXXXX1234123456}
                class=class="hl-string">"hl-comment"># Using current time as fallback if header parsing fails
                timestamp = datetime.now().isoformat()
                date_match = BLOCK_1_DATE.search(b1)
                if date_match:
                    class=class="hl-string">"hl-comment"># YYMMDD + HHMM
                    class=class="hl-string">"hl-comment"># Note: Sequence number often occupies this space, this regex is generic
                    class=class="hl-string">"hl-comment"># For simplicity, we assume file creation time or construct from header
                    pass 
                    
                class=class="hl-string">"hl-comment"># -- Extract Financials --
                amount = None
                currency = None
                val_date = None
                amt_match = TAG_32A_VAL.search(b4)
                if amt_match:
                    val_date_raw, currency, amt_str = amt_match.groups()
                    val_date = parse_swift_date(val_date_raw)
                    amount = float(amt_str.replace(class="hl-string">',', class="hl-string">'.'))

                class=class="hl-string">"hl-comment"># -- Extract Beneficiary --
                ben_bic = None
                ben_match = TAG_57A_BIC.search(b4)
                if ben_match:
                    ben_bic = ben_match.group(1)

                class=class="hl-string">"hl-comment"># -- Map Activity --
                activity = map_message_to_activity(io_dir, mt_type, b4)

                class=class="hl-string">"hl-comment"># -- Create Record --
                record = {
                    class="hl-string">"PaymentTransactionId": case_id,
                    class="hl-string">"ActivityName": activity,
                    class="hl-string">"EventTimestamp": datetime.now().strftime(class="hl-string">"%Y-%m-%d %H:%M:%S"), class=class="hl-string">"hl-comment"># Replace with actual extracted time
                    class="hl-string">"SourceSystem": class="hl-string">"SWIFT_Alliance_Access",
                    class="hl-string">"TransactionAmount": amount,
                    class="hl-string">"TransactionCurrency": currency,
                    class="hl-string">"UniqueEndToEndReference": uetr,
                    class="hl-string">"SwiftMessageType": fclass="hl-string">"MT{mt_type}",
                    class="hl-string">"BeneficiaryBic": ben_bic,
                    class="hl-string">"ValueDate": val_date,
                    class="hl-string">"LastDataUpdate": datetime.now().isoformat()
                }
                
                records.append(record)

                class=class="hl-string">"hl-comment"># -- Derived Logic for class="hl-string">'Created' --
                class=class="hl-string">"hl-comment"># If this is the very first outgoing MT103, we can imply creation
                class=class="hl-string">"hl-comment"># In a real scenario, this logic happens after sorting all events
                if activity == class="hl-string">"Payment Instruction Sent":
                    creation_record = record.copy()
                    creation_record[class="hl-string">"ActivityName"] = class="hl-string">"Payment Request Created"
                    class=class="hl-string">"hl-comment"># Offset time slightly backwards to simulate flow
                    creation_record[class="hl-string">"EventTimestamp"] = class="hl-string">"[Derive from Log Creation Time]"
                    records.append(creation_record)

        except Exception as e:
            print(fclass="hl-string">"Error processing file {filepath}: {e}")

    class=class="hl-string">"hl-comment"># -- Convert to DataFrame --
    df = pd.DataFrame(records)
    
    class=class="hl-string">"hl-comment"># -- Final Cleanup --
    class=class="hl-string">"hl-comment"># Sort by ID and Time to ensure logical order if needed
    if not df.empty:
        df.sort_values(by=[class="hl-string">'PaymentTransactionId', class="hl-string">'EventTimestamp'], inplace=True)
        df.to_csv(OUTPUT_FILE, index=False)
        print(fclass="hl-string">"Extraction complete. Data saved to {OUTPUT_FILE}")
    else:
        print(class="hl-string">"No valid SWIFT messages found.")

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

Your Payments Processing Data Template

Your Payments Processing Data Template

Payments Processing Attributes

Payments Processing Activities

Extraction Guides

SWIFT Alliance Access Archive Export

Steps

Configuration

a Sample Query config

Python Parser for Raw SWIFT MT Files

Steps

Configuration

a Sample Query python

SWIFT gpi API Connector

Steps

Configuration

a Sample Query python

Ready to Get Started?

Optimize Your SWIFT Payments Processing Now

Your Payments Processing Data Template

Your Payments Processing Data Template

Payments Processing Attributes

Payments Processing Activities

Extraction Guides

SWIFT Alliance Access Archive Export

Steps

Configuration

a Sample Query config

Python Parser for Raw SWIFT MT Files

Steps

Configuration

a Sample Query python

SWIFT gpi API Connector

Steps

Configuration

a Sample Query python

We value your privacy

Ready to Get Started?

Optimize Your SWIFT Payments Processing Now