Trading robots, built by traders.
TradeMachine AI was founded in 2023 by two traders and a machine‐learning engineer. The idea was simple: most people do not have the time to watch charts, read the news and follow the markets all day, so build robots that do it for them.
The first robot was built in 2023 on fifteen years of price data, news and calendar releases. It read a handful of indicators and traded a few currency pairs for a small group of beta clients. What they valued most was not having to do anything: they checked their account in the evening and got on with their day.
Since then the robots have learned to read more than 20 indicators at once, the financial news wires, the economic calendar and the posts of influential people on X. Each robot works within a budget and a daily loss limit, a lesson from the beta that has been on by default ever since.
Today there are 140 of us, about half in engineering and research, serving clients in 60 countries. Three robots run on our own hardware, every client has an account manager to call, and client money is held in segregated accounts.
- Founded
- 2023
- Clients
- 118,400
- Countries
- 60
- People
- 140
- Client money held
- $412M
- Robots
- 3
As of 31 Aug 2026. Client money figure from the most recent monthly reconciliation.
Four commitments, written into the product.
- Plain language. If a sentence needs the glossary, we rewrite the sentence. The client agreement is eleven pages.
- Show every trade. Every trade a robot makes is on your dashboard, with why it was opened and how it closed, losses included.
- Limits on by default. Every robot starts with a budget and a daily loss limit. Changing them is your decision, made with the numbers in front of you.
- People answer. Every client has an account manager, and chat, email and phone are staffed 24/7. The robots trade; people look after you.
Open roles
Engineering and research in Luxembourg and remote across Europe. Client support in Luxembourg, shifts covering 24/7.