High-Speed ISO 20022 Pacs.008 Payment Payload Repair & Routing System
In high-throughput interbank payment networks, even tiny syntactic errors in incoming ISO 20022 `pacs.008` messages (such as missing BIC codes, malformed IBANs, or corrupted postal strings) cause straight-through processing (STP) failures, forcing manual queue repair. In this challenge, you will implement an automated payment payload repair agent using CrewAI. The multi-agent swarm will collaborate to validate message fields, repair BICs via lookup registries, reformat malformed clearing tags, and re-inject validated ISO 20022 XML messages back into the STP rail.
What you are building
The core problem, expected build, and operating context for this challenge.
Develop a multi-agent ISO 20022 message repair swarm using CrewAI to restore high Straight-Through Processing (STP) rates.
How work is evaluated
Evaluates message repair accuracy and final ISO 20022 XML schema validity.
Challenge datasets
Public datasets, plus the private datasets you have access to.
How submissions are scored
These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.
xml_well_formed
Verifies output is well-formed XML and contains valid BIC code format
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
stp_repair_success_rate
Percentage of corrupted payloads successfully converted to valid STP state • target: 96 • range: 0-100
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
What you should walk away with
Configure specialized CrewAI agents (`xml_parser_agent`, `bic_lookup_agent`, `stp_validator_agent`)
Build custom BIC and IBAN lookup tools to repair missing ISO 20022 header elements
Maintain high Straight-Through Processing (STP) execution speeds without corrupting message payloads
Output valid, schema-compliant `pacs.008` XML strings
Reference links and supporting material
Synthetic set of corrupted ISO 20022 XML customer credit transfer messages containing field formatting errors.
Challenge setup
Evaluates message repair accuracy and final ISO 20022 XML schema validity.
Challenge input
Corrupted ISO 20022 XML payload string
CrewAI
Framework for multi-agent collaborative task execution.
Evaluated output
JSON specifying repaired_xml, bic_fixed, and stp_ready
- Verifies output is well-formed XML and contains valid BIC code format
- Percentage of corrupted payloads successfully converted to valid STP state • target: 96 • range: 0-100
- Benchmark: FinPayments-PayloadRepair-Bench
- Stp Repair Success Rate target: 96
- 1 public reference case
- Python execution harness
- Python sandbox (unavailable on Versalist)
Summary shows the first task, up to three tools, and up to three evidence items. Open the technical recipe to view all configured items.
View technical recipe
Configured tools
- CrewAI · Required
- crewAI · Optional
Evaluation contract
- xml_well_formed · Weight 1
- stp_repair_success_rate · Weight 1
Metrics
- stp_repair_success_rate · Target 96
Benchmarks
- FinPayments-PayloadRepair-Bench
- general
Reference cases
- 1 public reference case
Workflow tasks
- Task 1
Input: Corrupted ISO 20022 XML payload string
Output: JSON specifying repaired_xml, bic_fixed, and stp_ready
Execution harness: Python
Evaluation tools
- CrewAI · Agent Framework
Recipe freeze marker
No recipe freeze time is recorded.