The transparent, self-hosted AI analyst for finance teams
GenXFintel answers on your own data, in your own environment — every claim cited, every agent step visible, every dollar of compute accounted for.
Three kinds of data, one analyst
GenXFintel combines public market intelligence, your internal systems, and your uploaded documents — then reasons across all of them at once.
External market intelligence
SEC EDGAR filings and XBRL company facts, market-data APIs, news, earnings-call transcripts, and alternative data — normalized into one corpus.
Internal corporate data
Connect your own SQL warehouse and REST services so the analyst reasons over the numbers your business actually runs on.
Your uploaded documents
Drop in CSV, Excel, and PDF statements. They are parsed, chunked, and indexed alongside everything else — fully citable.
From filings to finished models
Two surfaces — a structured analytics dashboard and a conversational chat — backed by the same agent engine, tools, and data layer.
Multi-agent analysis
A Manager agent plans, a Delegator schedules, and Worker agents execute with specialized tools — every step visible live as it runs.
Quantitative workbench
Black-Scholes and Greeks, Monte-Carlo simulation (Heston, SABR), VaR/CVaR, and interactive 3D volatility surfaces.
Cited document grids
Ask one question across dozens of filings and get a grid of answers — every cell backed by a citation you can open and verify.
Editable deliverables
Export driver-based Excel models with live DCF formulas, and Word primers, SWOTs, and news briefs — not static screenshots.
Portfolio screening & backtests
Rule-based screeners, point-in-time backtests, efficient-frontier optimization, and holdings analytics.
Cost governance
Token usage and external API calls are tracked per analysis run, with budgets enforced — no surprise compute bills.
Ask. Watch the agents work. Verify everything.
Ask
Pose a question in chat or configure a structured analysis — a company deep-dive, a comparison, a valuation.
Agents plan & execute
Watch the Manager's plan, the task DAG, and each Worker's tool calls stream live. Nothing happens in a black box.
Cited, auditable results
Every number traces to a source line item or document chunk. Every run reports exactly what it cost.
Built for teams that can't outsource trust
Transparent reasoning
Live plan, task, and tool-call visibility for every analysis — auditable end to end.
Self-hosted & private
Bring your own data and run it in your own environment. Your filings, models, and questions never leave it.
Budgets, enforced
Per-run token and API cost tracking with hard budget limits.
Any LLM provider
Works with any OpenAI-compatible endpoint — Azure OpenAI, OpenRouter, or a local vLLM deployment.