About
Python backend, fullstack and AI systems
Full-Stack Developer · Python Backend Engineer · AI Systems Developer. I design and build server-side systems, interfaces and smart services for searching and processing information.
Mikhail Fediunin31 years old
Individual entrepreneur (IP), registered in Armenia

The core is a Python backend: APIs with FastAPI, Django and Flask, asynchronous programming, integrations between systems, queues and messaging. On the frontend — React and Next.js: multi-page websites, interfaces for the backend and admin panels.
A separate direction is AI systems and data: RAG search over knowledge bases, work with LLMs and vector stores, numerical processing, interpolation and data visualisation.
I use AI tools — OpenAI Codex and Claude — in my work, but architectural decisions, checking the result and responsibility for the code stay with me.
Technical profile
Backend
- Python
- Django
- FastAPI
- Flask
- REST API
- Asynchronous programming
- Integrations between systems
- Services and automation
Frontend
- JavaScript / TypeScript
- React
- Next.js
- Multi-page websites
- Interfaces for the backend
- Client applications and admin panels
AI / Data Engineering
- RAG
- LLM
- Intelligent search
- NumPy
- SciPy
- Pandas
- Matplotlib
- Numerical methods and interpolation
- Analytics and visualisation
Databases
- PostgreSQL
- pgvector
- Milvus
- Redis
- Cloud databases
- Vector search
Messaging and distributed systems
- RabbitMQ
- Apache Kafka
- Redis
- Asynchronous task processing
- Event-driven architecture
Networking
- Python sockets
- TCP
- UDP
- Client-server architecture
- Corporate chats
Automation and tools
- Python automation
- Bots
- API integration
- Docker
- Cloud infrastructure
- OpenAI Codex
- Claude
Also
- C#
- Object-oriented programming
- Software design
Approach
How I work
- 01
Architecture first
I start from the problem, the data and the load, and only then choose the tools.
- 02
Failure is part of the design
Error handling, timeouts and logging go in from the start, not after the first incident.
- 03
Simplicity pays off
Every abstraction has to earn its place. Separate services only when they solve a real problem.
- 04
Code for the team
Types, tests and a clear structure keep a project easy to change.
Have a system to build?
Tell me about the problem — I'll come back with questions and a proposed approach.