Daniel

Daniel Balzhyk
} https://t.me/somakash (preferred) # [open contact info](look above in the "contact info" section) ï [open contact info](look above in the "contact info" section)

Summary
Backend engineer focused on reliable async services and data pipelines. I design APIs and event-driven flows that
survive bursty traffic without drama. Comfortable with modern LLM tooling: retrieval, simple agent graphs,
MCP-style tool interfaces.

Work Experience
Cyclad Sep 2023 – Nov 2025 (2 years 3 months)
Backend Developer Warsaw, Poland (Remote)
• Shipped multi-tenant AI features: explicit tenant context in FastAPI, per-tenant vector indexes (pgvector/FAISS),
hybrid retrieval with citations, caching and token budgets. Exposed MCP-style tools (schema-first, auth, rate
limits) so agents could call internal actions safely.
• Built quick PoC LLM flows: compared GPT / Claude / Mistral / Gemini and different retrieval setups (dense vs
BM25 vs hybrid) on real client datasets. The PoCs we liked most were hardened into production flows with basic
observability and cost limits.
• Rebuilt event/postback pipeline on FastAPI + Kafka with idempotency, retry/backoff, DLQ and alerting. Result:
seconds-level delivery under peak, 5xx down by +-30%, fewer support escalations.
• Standardized offer onboarding via explicit contracts, schema validation and CI checks. Result: new providers
stopped blocking core development and were plugged in via the same CI-checked flow.
• Implemented multi-offer router with A/B and geo-fallback. Result: stable conversion across markets during
provider issues; smoother traffic utilization.
• Launched self-serve analytics with LLM/RAG (NL → SQL over warehouse). Result: PMs could answer most
routine questions themselves instead of pinging developers.
• Migrated services to Kubernetes (Helm charts, liveness/readiness, resource limits, HPA).
Stack: Python, FastAPI, SQLAlchemy, LangChain, PostgreSQL (RDS), Redis, pgvector, Kafka, Celery, Docker, Kubernetes,
GitLab CI, Nginx, ELK, Prometheus, Grafana, AWS S3, MongoDB, FAISS, RAG, MCP, CloudWatch, EKS, SQS
uWeb Agency Oct 2021 – Aug 2023 (2 years)
Junior Backend Developer Odesa, Ukraine
• Built a Django-based PDF reporting pipeline with templating and background jobs; sales got consistent,
ready-to-send documents without manual assembly.
• Delivered a reviews/ratings module with an admin moderation panel; user content stayed clean and trustworthy.
• Optimized heavy SQL and introduced Redis caching for high-traffic views; pages stayed responsive at peak while
DB load remained predictable.
• Standardized Docker environments for local dev and CI with Compose profiles and Make targets
• Added pragmatic tests on critical paths (pytest) and pre-commit checks (Ruff, mypy); fewer regressions and shorter
review cycles.
Stack: Python, Django, Django REST Framework (DRF), PostgreSQL, Redis, Celery, Docker, Nginx, Git, Pytest.

Technical Skills
Python, FastAPI, Django/DRF; asyncio, Pydantic v2; OpenAPI/Swagger, OAuth2/JWT, WebSockets, gRPC;
PostgreSQL, SQLAlchemy 2.0, Redis, Kafka, RabbitMQ, Docker, Kubernetes, Helm, GitLab CI/GitHub Actions,
Nginx, Linux; pytest, mypy, Ruff, Prometheus, Grafana, ELK, Sentry; RAG, LangChain, LangGraph, MCP,
Numpy/Pandas, GCP/AWS, OpenAI, LLM, CloudWatch, Lambdas, RDS, ECS, AWS Bedrock

Languages
English — Upper-Intermediate (B2) | Ukrainian — Native

Education
Odesa Polytechnic National University 2023
B.Sc. in Software Engineering (122) Odesa, Ukraine

1