About Termondt Tech

Fourteen years of trading infrastructure, applied to systems that run automatically and stop on command.

Termondt Tech is my engineering practice. I build the infrastructure behind a systematic trading operation: market data, decision support, automated execution and risk controls. Since 2026 I have been building it by guiding AI coding agents, under rules they have to obey and reviews they have to pass.

Based in Geneva, IL. The standard is the one I learned on exchange gateways and trading-firm platforms: infrastructure that has to be right, because real capital runs on it, and that a person can see into and stop.

What I am trying to do

Build technology precise enough to run a live trading operation automatically and transparent enough that a human can monitor and stop it. And show, in public and with the numbers, that an AI agent can write that kind of system safely when the discipline around it is right.

What the work involves

Right now, four things.

Data and decision support

Taking in live market data and turning it into ranked, explained decisions, with every one written to an audit log.

Execution and risk controls

Placing, watching and closing trades by rule, inside limits that are enforced twice and a kill switch that works from a phone.

Monitoring and reliability

Reconciling against the broker before anything resumes, alerting when something is wrong, and halting when the system can no longer prove what it holds.

Rails Reviews

Two weeks reading a team's agent prompts, their output and their tests, and handing back the written rules and pinned invariants the system is missing.

How I think about technology

Start with the failure modes

Before any code, work out what can go wrong and what the system does when it does. Then write the rule down with its reason, because a rule without its reason gets relitigated by the next person, or the next session, that finds it inconvenient.

Make the rule a mechanism

A rule that depends on anyone remembering it will be broken on an ordinary Tuesday. If the consequence is money, enforce it in code: a file lock, a required argument, a test that reads the source and fails if the forbidden thing exists anywhere.

No complexity for its own sake

The goal is not an impressive architecture. It is a system that does exactly what it needs to, that a person can read, and nothing more. One process, one event loop, one place that can send an order.

What sets this apart

✅

Fourteen years in trading infrastructure

Order-entry and market-maker gateways at the New York Stock Exchange. Market connectivity, an FPGA squad and the DevOps team at Chicago Trading Company. Monolith to services at Envestnet.

✅

Guided AI, on the record

Argus: 435 commits, 3,540 tests and 87 decision records in fourteen weeks, written by an AI agent under my direction and review, and run against a real brokerage account. The platform is public on GitHub.

✅

Automated and watched

Fully automated systems that a human can monitor and stop, in one action, from a phone.

✅

Evidence over spin

The first strategy lost money and was stopped on a test whose rules were written before the data arrived. The write-up says so, with the numbers.

Who I am

Rich Termondt
Founder and Principal

I have been building software since the late nineties, fourteen of those years in trading infrastructure. Eight years at the New York Stock Exchange on the order-entry and market-maker gateways, ending as Director of Trading System Development. Three at Envestnet as a principal architect, taking a monolith to services. Six at Chicago Trading Company, where I managed trading-technology squads including an FPGA team, founded the firm's DevOps team, and ran market connectivity for a stretch. For the back half of that career my job was managing engineers more than writing code: design reviews, code reviews, release sign-offs, retrospectives.

I founded Termondt Tech in 2024. In 2026 I ran an experiment on myself: build a platform for automated trading strategies, trade real money with it, and write none of the code. An AI agent wrote it, under rules I set and reviews I ran, and its first strategy was shut down on the evidence. The platform itself is public on GitHub, under an open-source license. Part one tells how the platform was built and what it took to make that safe. Part two tells how I knew the strategy had no edge, and how I knew I was not fooling myself.

Get in touch

If your team is putting AI-written code into production, or you run a trading operation that needs execution, risk controls or reconciliation built to run automatically, use the contact form and tell me what you are running.

Contact me