Ivan Poda

Backend & AI / Computer Vision Developer

Antalya, Türkiye

About

Backend & AI developer with 4+ years building real products - self-taught. I focus on AI integration, computer vision, and generative bots. Built real-time CV pipelines on live camera feeds (face-recognition attendance, automated counting, presence tracking), an AI custom-apparel marketplace with automated Printify fulfilment, and AI content-generation bots. I like owning the whole thing - backend, integrations, generation pipelines, deploy. Stack: Python, TypeScript, JavaScript, Node.js, React. AI/CV: OpenCV, RTSP, OpenAI API. Data: PostgreSQL, MongoDB. Infra: Docker, Linux.

Experience

  • Full-Stack Developer at FastyAds
    Feb 2026 - Present · 5 mos

    Building an AI content-generation platform for e-commerce (UGC photos, marketing videos, SEO copy) - solo, full-stack Own the entire product: frontend, backend, database, payments, and admin panel Designed token-based billing, subscription plans, and a multi-step generation pipeline

  • Full Stack Engineer at Mixora
    Sep 2025 - Present · 10 mos

    Built a custom-apparel marketplace end-to-end as the solo full-stack developer In-browser customizer for t-shirts, hoodies, and sweatshirts Automated order flow to Printify for on-demand printing and fulfilment Designed the backend, database schema, and frontend from scratch

  • Back End Developer at KanGram (Consulting)
    Apr 2024 - Present · 2 yrs 3 mos

    Developed a task-management system for Telegram groups Designed backend architecture (TypeScript/Node.js, PostgreSQL, TypeORM) Implemented a role-based access control (RBAC) system Deployed via Docker + CI/CD

  • Independent Developer at Freelance | Self-Employed
    Jan 2021 - Present · 5 yrs 6 mos

    Built AI content-generation Telegram bots (images and video from user prompts) Created a data pipeline that converted 3M+ PDFs of varying layouts into clean Markdown Shipped two Roblox games end-to-end (build, publish, marketing)

  • Computer Vision Engineer at Aivilon
    May 2025 - Aug 2025 · 4 mos

    Built real-time computer-vision pipelines on clients' live camera feeds (RTSP streams) Face-recognition system for staff attendance - automatic clock-in logging Automated product counting on the production line via object detection Presence tracking that flagged when staff left frame beyond set limits and pushed alerts