01
Modelled the research task, evidence levels, report structure, and analyst approval boundary.
Case study · Delivery evidence
Credit research automation without hiding the source
An agentic research workflow that connects internal materials and public financial information, exposes conflicts and missing evidence, and keeps investment judgement with the analyst.
By Onyx Devs Lab · Published · Updated
The operating challenge
Analysts repeatedly collected, reconciled, and reformatted information across internal files and public sources. Faster drafting was useful only if material facts, conflicts, freshness, and provenance remained visible.
Designed intervention
02
Connected internal materials with current public financial information through a deduplicated retrieval pipeline.
03
Separated facts, inference, and risk judgement while attaching sources to material conclusions.
04
Recorded analyst edits and missing evidence as evaluation data for future releases.
Validation snapshot
These figures separate delivered scope from pilot validation. Each metric retains its measurement definition so product coverage is not mistaken for business impact.
research source classes
Internal material and current public financial information
orchestrated stages
Retrieval, cross-checking, and report drafting
priority-fact recall
Reference facts entered the evidence set
conclusion citation coverage
Material report judgements retained a reviewable source
research first-draft cycle
Task creation to analyst-ready draft; 7.6 h before
Evidence boundary
The validation sample measures fact recall, citation coverage, and first-draft time. The system does not replace credit, investment, or trading decisions.