Search papers, labs, and topics across Lattice.
This paper introduces FinRCA-Bench, a synthetic benchmark designed to evaluate evidence retrieval and reasoning in financial AI systems, particularly for accounts-payable-to-bank reconciliation tasks. The benchmark consists of 2,250 cases with hidden root-cause labels and evidence contracts, allowing for an independent assessment of retrieval efficacy versus reasoning accuracy. Findings reveal that while classical machine learning achieves high accuracy, structural retrieval failures significantly impact overall performance, emphasizing the critical role of retrieval architecture in AI system effectiveness.
Retrieval architecture can dramatically influence AI performance, with structural failures outnumbering reasoning errors by a staggering 95 to 15 in financial reconciliation tasks.
Large language models are increasingly used to support financial operations, but their apparent reasoning performance can depend on whether they receive the right evidence. In financial reconciliation, the evidence needed for diagnosis is distributed across invoices, purchase orders, approvals, allocations, payments, ledger entries, and bank activity, linked by transactional relationships rather than textual similarity. End-to-end accuracy can therefore conflate evidence access with reasoning quality. We introduce FinRCA-Bench, a deterministic synthetic benchmark of 2,250 accounts-payable-to-bank reconciliation cases spanning 14 operational tables, including 1,500 injected failures across 15 causal categories and 750 legitimate or hard-negative cases. Root-cause labels and record-level evidence contracts are hidden from the model, allowing retrieval to be evaluated independently of answer correctness. We compare Rules/SQL, classical machine learning, dense semantic retrieval, deterministic relational expansion, and Typed Provenance Graph Retrieval (TPGR), a typed traversal restricted to persisted transaction relationships. Rules/SQL reaches 84.97% held-out exact accuracy and classical ML reaches 95.44%. Holding the reasoning model, prompt, and generation settings fixed while changing only retrieval increases macro required-record recall from 0.83% to 77.70% and exact 16-class accuracy from 2.05% to 72.44%. Structural retrieval failures outnumber reasoning failures with sufficient retrieval by 95 to 15; 254 correct predictions occur despite incomplete retrieval, and strict returned-evidence contract accuracy is only 5.72%. On FinRCA-Bench, retrieval architecture strongly shapes observed AI-system performance, and a correct root-cause label is a weak proxy for an auditable diagnosis.
