Python Developer / Backend Engineer with a background in computational physics and automation. I build practical, testable Python systems (FastAPI, MongoDB, CI/CD) and user-facing AI workflows (Streamlit + OpenAI API).
RP-first concept → playable build draft (SRD-only by default).
- Streamlit chat app with prompt guardrails and SRD-only constraints (public-safe).
- Optional Homebrew via UI toggle (homebrew is explicitly labeled).
- Session persistence (local) + test suite (unit/contract).
Repo: https://github.com/dmtr-karan/dnd-concept-to-build
SRD ingestion + validation pipeline for MongoDB, with CI artifacts.
- JSON Schema validation + idempotent ingest patterns.
- GitHub Actions pipelines for validation and artifact generation.
- Optional SRD “grounding” API for downstream apps.
Repo: https://github.com/dmtr-karan/dnd-srd-mongo
REST backend demonstrating auth + clean API design.
- JWT auth flow (
/register,/token) + protected routes. - Pydantic models and structured validation. Repo: https://github.com/dmtr-karan/fastapi-mini-blog
Interactive OpenAI-powered interview practice app (streaming, stop control, feedback flow).
Repo: https://github.com/dmtr-karan/streamlit-interview-simulator
- Python backend: FastAPI, Pydantic, REST APIs, MongoDB, testing (pytest), CI/CD (GitHub Actions)
- AI workflows: OpenAI API, prompt design/guardrails, Streamlit state & UX
- Automation background: experiment/data workflows, reproducible pipelines