Daria Chuprina

AI Engineer. Obsessed with soldering and programming gadgets out of literal garbage. Building things that actually work.

Based in Wrocław. I build AI-powered tools for European B2B companies — GDPR-compliant, production-grade, no bullshit architectures.

Currently deep into AI agents, MCP tools, RAG pipelines, computer vision, embedded systems, and learning ML — because apparently building one thing at a time is not an option.

Mom of two. Mass-producing code and humans with equal enthusiasm.


Python
C++
OpenCV
YOLO
MediaPipe
FastAPI
Django
React
PostgreSQL
Qdrant
Cohere
Azure
Docker
Celery
Redis
FastMCP


> pinned

Project About
🎩 jeeves Self-hosted multi-tenant AI assistant — RAG, 9 MCP tools, 5 messaging channels. Django 5 + React 19. ELv2.
🪟 nadya-vitrazh Interactive stained-glass installation — YOLO person detection + MediaPipe pose + state machine. HF Space — live test
🐭 dormouse Ukrainian text optimizer for LLMs — 73% token savings, 150% quality. pip install dormouse-ua PyPI
🔧 Vibecode-Cleaner-Fartrun Rust-powered code scanner with 29 MCP tools. Not another AI reviewing AI. Listed on mcpservers.org
🗓 mcp-ukrainian-calendar MCP server for Ukrainian pagan calendar — holidays, deities, folk wisdom. Listed on mcpservers.org

DOU — Tech Article of the Month Як я зекономила 50-73% на GPT API, бо в мене комп зі смітника — про dormouse-ua, написаний на Phenom II 2009 року.


> whoami

class Daria:
    location  = "Wrocław, PL"
    role      = "AI Engineer"
    kids      = 2
    coffee    = float("inf")

    focus     = ["AI Agents", "MCP Tools", "RAG Pipelines",
                 "Computer Vision", "Embedded Systems"]
    learning  = ["ML (the math, not just the API calls)"]
    passion   = ["Linux Kernel", "Raspberry Pi", "ESP32"]
    hero      = "Linus Torvalds"  # unironic

    addiction = "sci-fi"  # mass effect > sleep
    spirit_animal = "Sheldon Cooper"  # BAZINGA!

    @property
    def available_hours(self):
        return 24 - self.kids * 3  # optimistic estimate

> currently

🧠 Learning ML properly (not just calling APIs)
🛠  Building MCP tools, AI agents, and CV pipelines
👁  Person detection + pose estimation (YOLO + MediaPipe)
🔌 Embedded: ESP32, Raspberry Pi, GPIO, sensors
🐭 Published dormouse-ua — Ukrainian NLP for LLMs
📚 Reading more papers than a PhD student with a deadline

«Бо треба не всякою х*йнею займатися, а жити інтєрєсно: читати книжки, бути шпіоном, дресирувати любіму обізьяну»

— Лесь Подерев’янський


LinkedIn · GitHub · Threads · Reddit · Email

v0.3.3[beta]