About FinPile
Institution-grade stock analysis, written for regular people. Every stock gets its own AI agent, grounded in that company's filings, share counts, seasonality and ownership data — not the open web.
Who it is for
FinPile is an honest stock-analysis tool built for retail investors who want real data — not hype, not fabricated confidence, not dressed-up guesses.
What FinPile does
Type any ticker — stock or ETF, US, UK, or Canada — and FinPile runs six independent analyses and shows you each one honestly:
Fundamental
Financial health scored across 11 signals from real filings, with the score, the failures, and every caveat on the badge. Plus valuation against sector medians and the stock's own 5-year history — including the price band where our own verdict would change, and a DCF sensitivity grid that shows how violently “intrinsic value” moves when assumptions breathe.
Technical
A price ladder of the monthly and weekly levels the market has actually respected, unfilled gaps, trend structure, seasonal history over 10 years, and gap-behavior statistics — every rate shown as “X of N,” never as a prediction.
Sentiment
Reddit, X, StockTwits, and financial news aggregated and labelled with source counts and timestamps. Chatter is chatter; we tell you where the crowd leans and never pretend it's a signal.
Risk, Red Flags & Dilution
SEC filing history, share-count trends, offering and convertible detection, cash-runway pressure, jurisdiction facts, and the red-flag patterns behind most retail losses — pulled from EDGAR, scored in code.
Institutions, Insiders & Analysts
Who actually holds the stock (13F institutions with filing dates and passive tags, insider holdings derived from Form 4s) and how positions changed — labelled as the positioning history it is, never dressed up as “smart money signals.” Plus Wall Street consensus, presented without spin.
Catalysts, News & Earnings
What actually happened, classified by type and materiality. We retired “positive/negative catalyst” labels on purpose: whether news is good for shareholders is the market's call, not a classifier's.
Halal Screening
AAOIFI Standard 21 screening with a multi-standard comparison (Dow Jones Islamic, FTSE Shariah, MSCI), purification estimates, and the honest note when standards disagree — so Muslim investors see the methodology, not just a verdict.
Plus: a daily pre-market brief delivered by email, watchlist alerts when price reaches the levels that matter, and full ETF analysis — holdings, flows, and premium-to-NAV, computed only when the data is actually current.
How verdicts are produced
Every verdict in FinPile is computed in code from a named data source with fixed thresholds — the same rules for every stock, every day. When we can't verify something, we say so: you'll see “not meaningful” with a reason before you'll ever see a manufactured number. Our AI explains and summarizes what's already computed; it is never allowed to invent a figure, a rating, or a catalyst.
Every alert comes from the same deterministic, published methodology as every FinPile score, and every rendering states the snapshot's own as-of date.
And when the numbers can't honestly speak — a stock priced far beyond its reported results, a fund whose NAV is stale — FinPile says that, in a sentence, instead of printing false precision.
The one word at the top of a stock page — Buy, Hold or Sell — is decided by a rule you can read on the Methodology page, from a five-point checklist and the measured odds for the stock's class, and every call is scored on the Ledger. No price objectives, no forecasts, no opinions dressed as numbers. We give you a decision you can audit and the evidence behind it, and trust you to think.
Who builds it
FinPile is built by traders and investors who use it at their own desks, together with a set of AI agents built to work only from real data — filings, share counts, prices and disclosures — and never from their own guesses.
It started as a working problem: every retail tool either shouted trade signals it couldn't defend or buried the two questions that matter — is this business real, and what am I paying for it? — under hype. So the team built the tool it wanted: fixed rules, named sources, sample sizes in plain sight, and a hard ban on forward-looking claims.
The agents explain; code decides. Every number an agent cites is traced back to FinPile's own data before it is shown, and an answer that cannot be traced is withheld rather than shown. Every feature ships only after its arithmetic is checked by hand.