A complete AI research operation

for fund managers.

Finn monitors your book on its own, does the research you send it, verifies every number before delivery, and remembers everything. Personalized to your portfolio, your theses and your standards. Runs from one setup email.

A real Finn brief in the inbox, finding first with report and workbook attached

WHAT IT DOES

Seven jobs, shown one by one.

Tags show how each runs. Autonomous work happens without you asking.

Daily brief mock to supply

AUTONOMOUS

Morning brief

Everything portfolio-relevant that happened, read against your thesis, in your inbox before the open. Names with nothing material get exactly one line.

Large price moves trigger their own alert, with the explanations Finn checked and rejected.

FORWARD ANYTHING

Newsletter and sell-side filter

The 200 newsletters and broker notes you cannot read. Forward them. What touches your book comes back to you, and the rest is filed into searchable memory that resurfaces when it becomes relevant.

You forward a 40-page note and ask what matters here. Back come two paragraphs with page references, including where the analyst disagrees with your thesis.

A Finn evidence table compiled from many newsletter and broker sources
Finn workspace in earnings-season mode, with events tracked and dated

AUTONOMOUS

Earnings and filings

Every print, call and filing on your names, read the day it lands, against the KPI history Finn maintains. In reporting season, pre-print notes land before each release with what would confirm or break your thesis.

Release, call and transcript are tracked as separate checkpoints. An open call is never reported as no news.

AUTONOMOUS

Thesis drift tracking

Your thesis on each name becomes checkpoints, and Finn tracks the evidence in both directions. A weakening case surfaces early, while the exit decision is still cheap.

Every position carries its maintained thesis, the KPIs watched, and the next catalyst. This is what each new event is checked against.

Finn portfolio view showing the maintained thesis, KPIs and next catalyst for each name
A chart from a Finn report comparing headline and organic growth, with the falsification line visible

ON REQUEST

Models and analysis

Scenarios, sensitivities and peer comps, delivered as reports with the workbook attached. Live Excel formulas, so every figure traces to its input. After a print, the workbook is updated and re-sent.

You ask what happens to earnings if margins recover 200 bps. Back comes the worked scenario and the workbook behind it.

ON REQUEST

Meeting prep

Put the meeting in your calendar and the briefing arrives ahead of it. Where the company stands against your thesis, and the questions worth pressing on.

The briefing is built from the same maintained file as your reports, so it agrees with what you were last told.

Briefing email mock to supply

Decision ledger mock to supply

COMPOUNDS

Decision record

Every buy, sell, trim and pass saved with your reasoning at the time, then graded against what actually happened. Closed positions are never deleted, so the exit reason and the lesson stay on file.

Ask later whether your thesis actually played out, or what you keep getting wrong, and get the answer from your own record.

HOW IT WORKS

One run, start to finish.

Everything above goes through the same pipeline. Nothing is delivered until the gates close.

1. Work arrives

Your emails and forwarded research. Calendar events you schedule. Market moves, prints and filings.

1. Work arrives

Your emails and forwarded research. Calendar events you schedule. Market moves, prints and filings.

2. Your book, in memory

Portfolio and watchlist. Your thesis on every name. KPIs and standards, held in durable files.

2. Your book, in memory

Portfolio and watchlist. Your thesis on every name. KPIs and standards, held in durable files.

2. Your book, in memory

Finn begins reading filings, earnings calls, news, management commentary, and any research you forward.

3. The run

Everything new is read against your book. Numbers are fetched live. Tables are built in Excel first.

3. The run

Everything new is read against your book. Numbers are fetched live. Tables are built in Excel first.

3. The run

Finn emails you when something happens that affects your portfolio, with the source, the analysis, and what it means for your thesis.

4. Quality gates

Automated checks on citations and tables. An independent AI review against the sources. If it cannot finish, the report is held.

5. Your inbox

Finding first, in the email body. Report and workbook attached. Quiet names get one line.

Try Finn on your own portfolio.

Try Finn on your own portfolio.

Try Finn on your own portfolio.

Send one email with your positions, your watchlist, and what you care about on each name. That is the whole setup. Free two-week trial on your actual portfolio.