Andrej A.

I build and fix software that ships - with tests, not promises. You get working, reviewed, merge-ready code from one engineer accountable for the result. What I do: - Build & fix AI features - LLM agents, RAG pipelines, Claude/OpenAI integrations, MCP servers. - Automations & integrations - APIs, webhooks, data pipelines, browser automation/scraping. - Code review, bug-fixes & test coverage - I find the root cause, fix it, and prove it with tests. How I work: AI-assisted for speed, but I personally own the architecture, review and tests - clean code that fits your repo's conventions, passes CI and lints clean, not a raw AI dump. 10+ years architecting software (fintech/payments: data models, state machines, integrations). Fully async and written - one person accountable, no hand-offs. I work straight from your spec or a reproducible issue with minimal back-and-forth. Recent open-source sample: a 14-test PR adding characterisation tests to a complex orchestration class (full suite green) - see my portfolio below.