I build production-grade backends, high-throughput automation pipelines, and AI-native applications. With a foundation in Economics & IT (MGIMO University) and 7+ years of engineering experience, I focus on building systems that are not just technically sound, but cost-efficient and scalable.
- AI Infrastructure & Vector Search: Building hybrid semantic caching engines using Redis HNSW, Ollama, and GCP Vertex AI to reduce LLM API latency (<50ms) and slash token costs.
- AI-Native Engineering: Leveraging modern AI toolchains (Cursor, Claude Code, Agent harnesses) to build end-to-end full stack web applications and automated workflows at 3-5x velocity.
- System & Game Design: Solo-creator of "1983" (GMTK Game Jam — Top 1.7% in Narrative out of 10,500+ global entries), demonstrating deep systems thinking and real-time state mechanics.
- Languages: Python, TypeScript, JavaScript, SQL, HTML/CSS
- Backend & APIs: FastAPI, Node.js, Express, REST APIs, WebSockets, Celery
- AI & Data: Redis Vector Search (HNSW), Ollama, GCP Vertex AI, PyTorch, OpenAI / Anthropic APIs, RAG Architectures
- Databases & Infra: PostgreSQL, Redis Stack, Docker, GCP (Cloud Run, Compute Engine), Linux, Git
| Project | Tech Stack | Description |
|---|---|---|
| Semantic Caching Layer | FastAPI, Redis VSS, Ollama, GCP | High-performance semantic cache with real-time glassmorphic developer dashboard & hybrid search. |
| Text-to-SQL Guardrails | Python, LLM, SQL | Enterprise NL-to-SQL query generation pipeline with strict schema validation & security guardrails. |
| Documentation Web Crawler | Python, BeautifulSoup, Async | High-speed asynchronous scraping and markdown converter for RAG knowledge bases. |
📫 Connect with me: LinkedIn • Itch.io • Email: timursurov@gmail.com