I build AI systems that work in production — not demos. Custom RAG pipelines, intent-routing agents, and workflow automations for startups and scale-ups across EU and UK. Full-stack developer with 4 years of experience shipping production-grade LLM-powered systems, automation workflows, and complex platforms — from architecture through to live deployment. Deep expertise in n8n, Bubble.io, RAG pipelines, and LLM integration. Experienced working directly with founders and enterprise clients across full product cycles, with full ownership and no hand-holding. Delivered 8+ production systems including a multi-step RAG agent with 30,000+ vector rows, a multi-tenant B2B ESG platform serving 27+ enterprise clients, and two-sided marketplace platforms. I work async, deliver on time, and flag problems before they become yours.

Timofey Borovsky

I build AI systems that work in production — not demos. Custom RAG pipelines, intent-routing agents, and workflow automations for startups and scale-ups across EU and UK. Full-stack developer with 4 years of experience shipping production-grade LLM-powered systems, automation workflows, and complex platforms — from architecture through to live deployment. Deep expertise in n8n, Bubble.io, RAG pipelines, and LLM integration. Experienced working directly with founders and enterprise clients across full product cycles, with full ownership and no hand-holding. Delivered 8+ production systems including a multi-step RAG agent with 30,000+ vector rows, a multi-tenant B2B ESG platform serving 27+ enterprise clients, and two-sided marketplace platforms. I work async, deliver on time, and flag problems before they become yours.

Available to hire

I build AI systems that work in production — not demos.

Custom RAG pipelines, intent-routing agents, and workflow
automations for startups and scale-ups across EU and UK.

Full-stack developer with 4 years of experience shipping production-grade LLM-powered systems, automation workflows, and complex platforms — from architecture through to live deployment. Deep expertise in n8n, Bubble.io, RAG pipelines, and LLM integration.

Experienced working directly with founders and enterprise clients across full product cycles, with full ownership and no hand-holding.

Delivered 8+ production systems including a multi-step RAG agent with 30,000+ vector rows, a multi-tenant B2B ESG platform serving 27+ enterprise clients, and two-sided marketplace platforms.

I work async, deliver on time, and flag problems before
they become yours.

See more

Experience Level

Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
See more

Language

English
Advanced
Polish
Fluent

Work Experience

AI Product Engineer / Low code Developer at WEQUITY
August 25, 2023 - July 1, 2025
Develop and design intuitive, user-friendly applications, and work closely with product, web, data, and design teams within a fast-paced, startup environment. Delivery high-quality solutions swiftly.
Low-Code / AI Product Engineer at Freelance
May 6, 2022 - Present
- Delivered 8+ production applications over 24 months across AI, marketplace, enterprise, and civic tech sectors. - Built a multi-step RAG system with 30,000+ vector rows, semantic retrieval, intent routing, and text/voice I/O — deployed to production for a Belgian retail client. - Built two-sided marketplace platforms with Stripe multi-party payments, custom booking engines, and dual-role user flows. - Built task management, trendwatching, and conversational AI tools with text and voice interfaces.

Education

Bachelor of Marketing at University of Lodz
October 1, 2022 - October 15, 2025

Qualifications

Million Labs Certified Bubble Associate
October 1, 2024 - April 10, 2026

Industry Experience

Software & Internet, Computers & Electronics, Wholesale & Distribution, Professional Services, Media & Entertainment, Energy & Utilities
    paper Custom RAG AI Agent — Retail Recommendation System (Belgium)

    Production RAG System with Intent Routing & Custom Architecture

    Built a multi-step AI agent for a Belgian retail client —
    not a generic LLM wrapper, but custom architecture optimised
    for low latency and high accuracy.

    How it works:

    • Router node classifies intent: discovery / general Q&A /
      recommendation — and structures chat history before routing
    • Three separate passes handle each intent type independently
    • RAG layer uses semantic retrieval + custom relevance scoring
      • strict filtering (not a simple vector search)
    • Parallel sub-workflows surface upsell products simultaneously
    • Validation agent checks output, corrects hallucinations,
      returns clean structured response

    Why it’s not a standard RAG:
    Most RAG systems connect an LLM to a vector DB and call it done.
    This system has custom scoring logic, intent-aware routing, and
    a validation layer — built for production reliability,
    not demos.

    Stack: n8n, Supabase, custom scripts / webhook architecture,
    semantic retrieval pipeline

    ⚠️ Client is under NDA

    paper Civic Election Web App — Belgium 2024 (Desktop + Mobile)

    Public Civic Tech Platform — Belgium Elections 2024

    Built responsive desktop + mobile web app helping Belgian voters
    compare candidates across 10+ political parties ahead of the
    2024 regional, federal, and European elections.

    My scope:

    • Full responsive + mobile implementation from scratch
    • Admin panel development
    • Custom API endpoint integration
    • WU optimization & performance improvements
    • Bug fixing + feature delivery before public beta

    Stack: Bubble.io, Custom APIs, JavaScript

    🔗 [Live platform](https://www.twine.net/signin

    paper Two-Sided Marketplace App — Cleaning Services (Netherlands)

    Localized Service Marketplace — Full Cycle Delivery

    Built a dual-sided marketplace connecting families
    with cleaning professionals across the Netherlands. Delivered
    full MVP (from discovery to first user) in under 6 weeks.

    Key features I built:

    • Custom calendar + booking flow (JavaScript)
    • Stripe integration with iDEAL (Dutch payment method)
    • Real-time in-app chat + notification system
    • Google Maps API with custom JS map layer
    • Availability management for service providers
    • Admin panel + privacy rules

    Stack: Bubble.io, Stripe, Google Maps API, Loops, JavaScript

    ⭐ 5.0 on Upwork — “Tsimafei exceeded my expectations.
    He knows exactly what needs to be done and implements features
    that enhance the success of your app. He is exceptional.”

    🔗 [Live app](https://www.twine.net/signin

    paper AI-Powered ESG SaaS Platform — 27+ Enterprise Clients

    ESG Compliance Platform with AI Automation

    Built and scaled a B2B SaaS platform that automates complex ESG
    questionnaires (EcoVadis, CSRD) using AI — reducing completion
    time by 70%.

    My scope:

    • Grew platform from 0 → 27+ paying corporate clients
    • Managed 500,000+ live database entries at 3s load speed
    • Completed 300+ tasks / 3,500+ hours on this project
    • Led feature development across the full product cycle

    Key systems I built:

    • AI document extraction & custom request processing
    • Knowledge base with multi-file upload + real-time storage monitoring
    • Team management: roles, permissions, workspaces, chat
    • Auth0, OpenAI, Stripe, Google Cloud, SendGrid integrations and much more
    • Version control, branch management, deployment

    Stack: Bubble.io, OpenAI API, Stripe, Auth0, Google Cloud,
    Zapier, SendGrid, LokaliseJS

    🔗 [Live platform](https://www.twine.net/signin