Skip to content
View sviat-barbutsa's full-sized avatar
3 followers · 2 following

Achievements

Achievement: Pair Extraordinairex4Achievement: QuickdrawAchievement: YOLOAchievement: Pull Sharkx3Achievement: Arctic Code Vault Contributor

Achievements

Achievement: Pair Extraordinairex4Achievement: QuickdrawAchievement: YOLOAchievement: Pull Sharkx3Achievement: Arctic Code Vault Contributor

Block or report sviat-barbutsa

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
sviat-barbutsa/README.md

Sviat Barbutsa

Senior Frontend Engineer / Frontend Architect — React · Next.js · TypeScript · AI Integration

13+ years building web platforms from the ground up. Currently Lead Software Engineer on an AI-powered collaborative SaaS: multiple React/Next.js apps and a shared component library in a single monorepo, on top of 15 Cloudflare Workers microservices. I do frontend systems and ship production AI features (RAG, vector search, LLM routing) by day and build local-first AI agents by night.

Featured projects

llamail — Your private email agent, running locally. Control Gmail from Telegram: LLM-as-router intent classification, hybrid RAG search (ChromaDB + SQLite FTS5), drafting, campaigns, scheduled sends. Runs on consumer hardware — zero cloud, zero monthly bills.

small-context-rag-solution — Python + TypeScript patterns for RAG under small context windows: summaries, raw chunks, and context-budget tracing. Companion code to my freeCodeCamp article.

ai-memory-tiers-demo — A dependency-free demo of a 3-tier memory system for AI coding agents that cut my coding context by 96%. Companion code to my Medium article.

Writing

Stack

Frontend: React, Next.js, TypeScript, React Native, Astro, Redux/MobX, GraphQL, tRPC, Tailwind, design systems, accessibility, i18n Backend & platform: Node.js (Nest.js, Hono), Python (Django, FastAPI), Cloudflare Workers / D1 / R2 / KV / Durable Objects, WebSockets, Docker, CI/CD AI engineering: OpenAI & Gemini APIs, RAG (Vectorize, AutoRAG, ChromaDB, SQLite FTS5), embeddings, LLM routing, local LLMs (llama.cpp), LangChain

Elsewhere

LinkedIn · StackOverflow (5k+ rep) · dev.to · Medium · sviatoslavbarbutsa@gmail.com

Pinned Loading

  1. llamail llamail Public

    Your private email agent, running locally. Control your inbox from Telegram - RAG search, drafts, campaigns. Runs on consumer hardware. Zero cloud, zero monthly bills

    Python

  2. small-context-rag-solution small-context-rag-solution Public

    Companion code to my freeCodeCamp article 'How to Handle Small Context Window Limits in RAG Systems' — Python + TypeScript patterns: summaries, raw chunks, context-budget tracing.

    TypeScript 1

  3. ai-memory-tiers-demo ai-memory-tiers-demo Public

    Companion code to my Medium article 'The 3-Tier Memory System That Cut My AI Coding Context by 96%' — a dependency-free demo.

    JavaScript

  4. cronix cronix Public

    Paper-trading terminal for KuCoin — FastAPI + React, bracket orders with draggable chart lines, JWT auth, risk calculator

    TypeScript

  5. agentic-flow agentic-flow Public

    Multi-user chat where an AI agent turns UI screenshots into Web Components — approval-gated builds, live previews, automatic Git branch delivery. Node built-ins only.

    JavaScript