Tom Pickel
CAIA, FDP
Judgement and engineering, together.
Pickel Fintech is my practice. Investing judgement and engineering live in one person here, so a question that starts in a conversation tends to end up as a tool, a report or a decision you can actually act on.
I approach markets as systems shaped by human perception. The work is an attempt to understand that dynamic and act within it, using the analytical discipline of institutional investing and the leverage of modern software.
How I got here
Twelve years ago I began with company valuations, business planning and financial models. That grew into analyzing alternative strategies (private equity, debt, real estate, infrastructure and others) for institutional investors, where I learned how capital thinks at scale.
Seven years ago I stepped into programming and data, first out of curiosity, then out of necessity. The problems I cared about had outgrown what a spreadsheet can do. The CAIA and FDP charters mark points along that path, within a pursuit that continues to evolve.
In 2021 I left my job to experiment more directly. The first project was a neural network built from scratch to model equities. Since then I have kept testing how data and software can extend investment insight. Some of that thinking generalises into products. Some stays specific, and becomes advisory. The two feed each other.
Products I build and run
Five products are live today. Each grew from a real question I wanted answered for my own work, then opened up for others. They are refined every week as the data teaches us more.
Most portfolios are collections of good ideas, not coherent systems. Souppe scores a portfolio across eight independent risk dimensions, shows where exposure is strong and where it is thin, and suggests the next security to add - one at a time, with measurable before-and-after impact.
SEC 13F filings carry billions of dollars of signal, buried in paperwork. The 13F makes that signal readable: positioning, crowding and conviction across 8,600+ institutional managers, from a single manager's deep-dive to market-wide capital flows. Built for investors who want to know what capital is actually doing, quarter by quarter.
Comment sections shape how readers experience a publication. SentiSift brings real-time moderation, analytical intelligence on reader discussions and active participation that lifts the quality of conversation, with multilingual coverage at 94% accuracy using an ensemble of 8+ AI models.
Markets move on events faster than anyone can track by hand: earnings, deals, filings, launches and shocks. Cuebit watches around the clock, merges duplicate reports of the same event into one clean record, matches it against your standing alerts and emails a sharp brief - what happened, why it matters and the source. Maximum signal, minimum words.
US House members report their stock trades, but the disclosures are scattered and slow to read. Capitol Gains checks the official filings several times a day and emails each newly reported trade, whether you follow every tracked member or only the top 100 ranked by risk-adjusted return against the S&P 500.
Advisory
Advisory is organised into two specialties: investment & decision work (asset valuation, alternative investment analysis, portfolio consultancy, investment decision memo, institutional positioning report) and data & engineering work (investment data systems, operational data & process systems, custom data & automation engineering). The same practice that powers the products, scoped for your specific situation. Investment judgement and engineering from one person, so the work moves cleanly from question to answer - whether about a portfolio, a financial decision, a system to build or a bespoke engineering need.
Writing at pickel.io
Alongside the products and advisory, I write at pickel.io. Writing is how I sharpen my thinking, surface gaps in my own knowledge and keep developing how I perceive markets. Much of what later becomes a product feature or a decision memo started there as a rough, half-formed idea. The record is open, including the parts I do not yet have right.
How the work actually feels
Most of it is quiet. Reading filings, scrubbing data, testing a hypothesis, shaping a model until it runs clean. The visible output (a product, a report, a decision memo) sits on top of a lot of craft underneath. That is the part I care about most.
If you are working on something hard and want a thinking partner who can also build it, a short conversation is the easiest way to start. The best of this work has come from people who brought a question I had not yet thought to ask.