Kepler, a verifiable AI platform for financial research, has expanded on the system architecture highlighted in Anthropic’s recent customer profile, detailing how it combines language models with deterministic financial software to deliver auditable and defensible financial analysis. Designed specifically for the rigor of investment research and compliance workflows, Kepler separates AI-driven reasoning from data retrieval and calculations. Language models interpret analyst queries and generate narratives, while code handles financial data extraction, calculations, and citations through a proprietary financial ontology that maps analyst terminology such as EBITDA and free cash flow to exact filing line items.
Also Read: Nuvo Technologies Integrates With Avalara to Unify AI-Powered B2B Onboarding and Tax Compliance
The platform indexes 26 million SEC filings, 50 million public documents, and 1 million private documents across 27 markets, ensuring every figure is traceable to its source. “In finance, the model can’t be the whole system,” said Dr. John McRaven, CTO of Kepler. “We treat it as one stage in a pipeline whose job is to hand the model exactly what it needs to succeed at exactly that stage.” CEO Vinoo Ganesh added, “On our workloads, Claude was the model that consistently held the plan together. Other models would start strong and then quietly drop a constraint by step five. That behavior matters more than any benchmark score. One wrong assumption early in a financial analysis breaks everything downstream.”
