Сергій

SERHII MUZYKA
Python Backend Engineer | LLM Agent Development
[open contact info](look above in the "contact info" section) [open contact info](look above in the "contact info" section) Ternopil, Ukraine
[open contact info](look above in the "contact info" section)

Summary
Python engineer combining production backend experience (async services, Docker, Prometheus/Grafana,
hardware-integrated systems) with hands-on work building LLM agent systems (tool orchestration, RAG, long-term
memory). Comfortable owning a scoped deliverable end-to-end and working against a task list and deadline rather
than clocked hours — available for contract, freelance, and milestone-based engagements.

Experience
Software Developer, Vodaland Industries 11/2024 – 07/2026 | Ternopil, Ukraine
Built and scaled a factory monitoring platform integrating 10+ cameras, 100+ sensors, and 6+ PLC controllers for
•

real-time fault detection.
Redesigned camera-based product measurement using Sobel edge detection and gradient analysis to locate part
•

edges directly, replacing a manual two-point sampling method and improving accuracy and false-positive rate.
Integrated 6 additional TCP-connected containers streaming data from mixers, silos, and an accelerometer used
•

to detect excessive vibration on critical equipment, storing metrics in Prometheus.
Built 30+ Grafana dashboards spanning equipment-online status, per-line state views, per-equipment-type
•

overviews, and single-machine operator control views.
Developed a Flask-based real-time dashboard and control panel for equipment tracking.
•

Designed async, fault-isolated services using Docker, Prometheus, and Grafana.
•

Selected Projects
LLM Agent Framework (“AI Copilot”) 09/2025 – Present
Personal R&D project - https://github.com/st000ne/Copilot
Built a lightweight local agent framework on top of raw LLM API calls, implementing tool-calling (web search,
•

translation) and a FAISS-backed long-term memory layer — giving the agent context persistence across multi-turn
conversations, distinct from a stateless API wrapper.
Implemented an async backend serving a React frontend for real-time interaction.
•

Job-Board Backend (LinkedIn-style platform), Personal project - https://github.com/st000ne/Job_Board
Designed and built a full backend for a job-board/social platform: JWT auth with role-based access control
•

(Admin / Company / Employee), async SQLAlchemy + PostgreSQL, Alembic migrations.
Implemented a semantic job-recommendation engine using sentence-transformer embeddings and FAISS
•

similarity search to match employees to relevant openings.
Added background task processing (Celery/Redis), rate limiting (SlowAPI), transactional email (SendGrid), and
•

automated tests (pytest, httpx, pytest-asyncio).
Spy Cat Agency, Full-Stack Technical Assessment - https://github.com/st000ne/SpyCatAgencyBackend
Delivered a complete FastAPI + Next.js/TypeScript CRUD application under a compressed deadline (under 6
•

hours): spy cat, mission, and target management with enforced business rules (1–3 targets per mission, single
active mission per cat, locked state after completion).
Integrated a third-party API (TheCatAPI) for breed validation with in-memory caching to reduce external calls.
•

Education
Bachelor's Degree in Automation and Computer-Integrated 09/2022 – 06/2026 | Ternopil, Ukraine
Technologies
West Ukrainian National University

Skills
AI & LLMs Backend
Agentic workflows (ReAct), LangChain, tool Python (async), FastAPI, Flask, SQLAlchemy,
orchestration, RAG, semantic search (FAISS), prompt TypeScript, Next.js, Celery, Redis
engineering
Tools & Workflow
Systems & Infra Git, CI/CD basics, Agile/Sprints, browser automation,
Docker, Prometheus, Grafana, Loki, PostgreSQL, REST Human-in-the-Loop AI design
APIs, PLC/TCP integration

Availability
Remote-first, open to short sprints or longer retainers. Comfortable with milestone/deliverable-based contracts.
Fluent English (C1).