Senior Full-Stack AI Product Engineer

Data Science UA is a service company with strong data science and AI expertise. Our journey began in 2016 with uniting top AI talents and organizing the first Data Science tech conference in Kyiv. Over the past 9 years, we have diligently fostered one of the largest Data Science & AI communities in Europe.

About the role:
We are looking for a Senior Full-Stack AI Product Engineer.
The archetype for this role is a founding engineer at an AI startup: someone who has personally taken an AI product from nothing to live users. We need an engineer with broad context rather than narrow depth, comfortable making architectural decisions independently, and focused on shipping fast and reliably.

Responsibilities:
- End-to-End AI Product Delivery: Architect, build, and ship a production-ready AI application from scratch to live users, taking full ownership of both front-end (React/Next.js) and back-end (Python/Node) development without relying on a designer or supporting team.
- Real-Time Voice & Video Integration: Implement and optimize real-time conversational voice/video APIs (e.g., Tavus, LiveKit, Pipecat, Vapi, OpenAI Realtime, Gemini Live) to ensure seamless, low-latency user interactions.
- Latency & Performance Optimization: Measure, troubleshoot, and optimize system response times in milliseconds to eliminate delays and maintain a highly responsive user experience.
- Document Processing & Retrieval (RAG): Build retrieval systems over complex document sets (PDFs) to ensure the AI generates accurate, source-cited responses.
- Deployment & Infrastructure Management: Deploy and manage application services on modern managed platforms
- Technical Decision-Making & Communication: Independently make core architectural choices and clearly explain technical trade-offs in plain language to non-technical stakeholders and legal team members.
- Sprint Ownership & Weekly Demos: Proactively drive project momentum, scope-manage feature requests to meet tight timelines, and demonstrate functional, working software on a weekly basis.

Requirements:
- Has shipped a production application built on large language models, end to end. Not a prototype, not a hackathon project - something real users used. They should be able to name it and describe what they personally built.
- Full-stack capability, unaided. Front end and back end. Typically React or Next.js with Python or Node. They must be able to build a clean, usable interface without a designer.
- Has integrated at least one real-time voice or conversational video API. Any of: Tavus, LiveKit, Pipecat, Vapi, Retell, ElevenLabs, Deepgram, Cartesia, OpenAI Realtime, Gemini Live. This is the single most important technical filter.
- Understands and has optimised latency. Can talk about response times in milliseconds and has made something faster. In this product a two-second delay ruins the illusion.
- Has been the sole or lead engineer on something that shipped. Evidence they can operate without a team around them.

Nice to have:
- Retrieval over document sets ("RAG") where the AI must cite sources accurately
- Experience with Tavus specifically — the client's existing prototype uses it
- Prior work at a seed-stage startup, or as a technical co-founder
- PDF and document processing
- Deployment on a managed platform (Vercel, Render, Railway, Fly.io)