Our first product, for investment professionals.
Find, screen, analyse documents and create reports in a single place.
It’s the foundation. Every answer, every figure, every summary traces back to a line in a document.
Our ingestion, OCR, and entity layer are our own. We control the primitives so the quality compounds. No shortcuts.
Analyst One behaves like a seasoned analyst. Not standing in your way, but delivering precise, skeptical, and accurate work.
Surface the most relevant research instantly across every source you cover.
Extract insights from filings, reports, and presentations in seconds — with citations.
Institutional-grade research reports. Full audit trails. Your data, new insights.
Enterprise-grade controls. Full audit trails. Your data, never ours.
Call Felix from Claude or Codex, hit it from the CLI, or wire it into your stack via API.
Share findings, annotate research, and move together — live, with your whole team on the same page.
Track companies and news, and pull in external data from high-quality providers — all in one place.
Analyst One lives inside the tools you already use. Call it from Claude, use it from Codex, wire it into your own pipeline with the API, or pull it up in a terminal with the CLI. The research layer follows you — not the other way around.
Roots in private equity, AI governance, alternative-asset trading, and defense-grade engineering. We’ve spent years inside the data frictions we’re now solving.
Our market update — the AI trade, macro, M&A and public markets read for signal, not noise, every figure sourced. Published on the site, free to read.

Short essays on the mechanics of research, market commentary, and notes from the team. New posts twice a month.
Still curious? Drop us a line — we read everything.
Our primary focus is the buy-side — credit analysts, private equity teams, and institutional researchers — but we don't judge. Anyone whose job involves reading dense financial documents and turning them into a defensible point of view can use Analyst One and see it in action.
Analyst One is the analyst bench we've built from the ground up — search, screening, document analysis, and citation in a single workspace. Build 1.0 is live today.
We treat documents as documents, not as flat text. Tables are parsed as tables, footnotes stay attached to their callouts, multi-column layouts are reconstructed, and structure is preserved through the pipeline rather than flattened. Our extraction layer was built in-house specifically for financial filings.
Analyst One is browser-first today, with native integrations into Claude, Codex, and a CLI/API surface arriving in H2 2026. Runs in any modern browser — Chrome, Safari, Firefox, Edge.
The parsed index lives on your device; search and citation run locally. Pages we process for OCR are auto-deleted on return, with a 24-hour hard ceiling. Model calls travel via zero-retention, no-training contracts. We're moving toward sealed inference where possible.
Analyst One is tiered by seat with volume discounts for teams of 10+. We'll publish the full schedule when general availability opens — in the meantime, talk to us and we'll walk through what fits.