GENIE X HUBPROJECT: ECLIPSE
BACK TO ALL CHALLENGES
02
CHALLENGE 02Finance

Autonomous Resolution of Cross-System Transaction Mismatches

Match and resolve transaction discrepancies across financial systems with an auditable AI agent.

Problem Context

Financial reconciliation teams spend enormous effort manually matching transactions across payment processors, ERP systems and bank statements. Duplicate payments, partial settlements and missing references create costly exceptions that require expert investigation.

Challenge Requirement

Build an AI agent that ingests transaction data from multiple systems, detects mismatches, retrieves supporting evidence, classifies discrepancy types, recommends resolution actions, and produces an auditable action trail — with human approval for consequential financial updates.

Mandatory Capabilities (8)

01Cross-system data ingestion
02Transaction / entity matching
03Evidence retrieval
04Context-aware classification
05Partial / duplicate / missing payment detection
06Action recommendation
07Auditable action trail
08Human approval for consequential updates

Illustrative Scenario

A mid-month reconciliation reveals 47 unmatched transactions between the payment gateway and the ERP. The agent cross-references order IDs, timestamps and amounts, identifies three categories — duplicate retry charges, currency rounding errors and missing settlement references — and proposes automated write-offs for sub-threshold differences while escalating consequential corrections for finance team approval.

Safety & Governance Rules

Consequential financial updates must be gated by human approval. The system must maintain an immutable audit log of all classification decisions and actions taken.

RECOMMENDED EXECUTION TRACE

Ingest multi-system transaction data
Plan matching strategy
Execute entity-matching tools
Classify discrepancy types
Retrieve supporting evidence
Generate action recommendations
Request human approval for consequential items
Log all decisions & actions

BUILD REQUIREMENT

Demonstrate an end-to-end reconciliation flow: ingest mismatched transaction data, classify each discrepancy, retrieve evidence, recommend actions, obtain human approval for consequential updates, and produce an auditable action trail.

End-to-End System Required

SELECT THIS TRACK

Register your team and select Challenge 02 as your primary track.

REGISTER NOW