About
My name is Ilko Vasilev and I am from Razgrad, Bulgaria. With the help of AI, I build tools for finance, analysis and automation. They collect data from many places, score it and run on their own. I taught myself through real projects, and I check everything before I call it finished.
In short
I turn a vague question into a working system. I describe what needs to be built, with what data, and how we will know that it works. Then I direct AI as it writes the code, test it on real data and send it back for fixes until the result passes the test.
I am not a classic programmer and I do not pretend to be one. My strength lies elsewhere: I understand the problem, I know what to ask for, and I do not accept a number I cannot verify. That is how I built the four projects in this portfolio, from a sentiment terminal with more than 100 sources to an analysis of 751,500 customer reviews.
Before that I worked in the warehouse of Ozone.bg and as a sales floor supervisor in a supermarket. That is where the habits I still use come from: accuracy under pressure, clear procedures and ownership of the result.
What I do well
I
Systems that work without me: they check every few minutes, keep a history in a database, send Telegram alerts and come with a dashboard in the browser or on the phone. The AI models run on an ordinary computer, with no paid services.
Proven in Market Sentiment Terminal and Strategy Terminal
II
Collecting data from many sources, cleaning it and removing duplicate and wrong records, then analysis with a clear conclusion. I check the method against independent labels, and every number in the text is matched automatically against the data.
Proven in the review analysis and the inventory plan
III
Since 2023 I have studied the financial markets on my own: technical and fundamental analysis of crypto, forex, stocks, ETFs and futures. I do not judge ideas by eye but with backtests, real fees and a comparison against random entries.
Proven in Strategy Terminal and Market Sentiment Terminal
The path
No university. Everything after high school I learned on the job and through projects.
2022
English and Russian profile. My English is at working proficiency.
After high school
Responsible for the shift, the tills and reconciling daily revenue, restocking, and solving customer issues on the spot.
Next
Processing orders for a high-volume online store: scanning, tracking stock and checking every parcel.
Since 2023
Technical and fundamental analysis, consistently and on my own. I wanted to test my ideas with data rather than gut feeling, and that led me to build my own tools.
Today
A sentiment terminal, a signal system with a backtest, a review analysis and an inventory plan. Four projects that work and have been tested.
Approach
I start from what needs to be solved and what a good result looks like. The technology comes after that.
Backtests with fees, comparisons against chance, human labels, automatic checks of every number. A pretty chart proves nothing.
If something does not beat chance or the data is not enough, I say so. That is how good decisions get made.
A working app, a server running 24/7, a report that downloads to Excel. Not a prototype that works only once.
What I can do
What my projects are built on
What I am looking for
Data analysis, fintech, AI automation, operations or support. I prefer remote work, and I am ready to relocate for the right role. I would be glad to show my projects live.