Who I am
Nadir Abdurakhmanov
Software engineer
I design and build software that does repetitive work instead of people. AI agents and bots, websites and web apps, backend and integrations, mobile apps, trading infrastructure. From the first call to production, one person owns the result — me.
No office, no managers, no sales department. Which means none of their salaries sit in your quote, and you talk directly to the person who will write the code. The honest downside: I take two or three projects at a time, and the queue is sometimes a month deep.
How it works
- I work alone
- For design and mobile I bring in people I have worked with for years
- 2–3 projects at a time
- So nothing stretches past its deadline
- You talk to me
- No manager relaying messages in between
- Two markets
- Europe, Russia and the CIS. Contract in euros or roubles
The proof
My own trading system in production
For several years I have been building and running my own trading system: eleven services, round-the-clock market data collection, order execution, position accounting, a risk layer with a kill switch. Hot-path computation lives in a Rust module, the rest is Python, PostgreSQL, Redis, Docker.
It is not a showcase, it is a bench. Every decision I offer clients was first paid for with my own mistakes: queues that accumulated tens of seconds of lag, position closing that silently never fired, a cleanup routine that deleted live data. So when I take on someone else’s system, I bring a concrete list of where these projects break rather than general reasoning.
What I decline
A short list of noes saves time on both sides. If your task falls into it, better to know now than after three meetings.
- —I decline work that is solved by configuring something you already pay for
- —I do not sell trading signals or manage other people’s money
- —I do not promise revenue growth — I answer for the system doing what the contract says
- —I never keep a project on my own servers: access is yours from day one