Skip to content
View kapustazh's full-sized avatar
💭
^-^
💭
^-^

Block or report kapustazh

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.

Content in all repositories owned by your account will be closed.
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
kapustazh/README.md

Matvii Nesterenko

Systems, Web3 & Software Engineer · Low-Level Systems · AI Agent Tooling · Graph Algorithms Heilbronn / Stuttgart, Germany · work@kapustazh.dev · kapustazh.dev · LinkedIn


Projects

Project Description Status / Stack
OneShot Agent-oriented Web3 system for reliable execution and settlement of approved intents. Built as a multi-component TypeScript application with API, worker and UI services, PostgreSQL persistence, MCP integration and automated tests. Active Development · TypeScript · PostgreSQL · MCP · Web3
DeepTrace Model Context Protocol (MCP) server delivering verifiable DeFi data on Base to AI agents. Enforces source attribution and freshness verification via NutHatch Indexer to eliminate LLM hallucinations. Features 4 agent tools for Uniswap V3 analytics and USDC lending rates. 🏆 3rd Place: Best AI Tooling ($1,000 The Graph Bounty) @ ETHGlobal Lisbon 2026 · TypeScript · MCP · Base
call_me_maybe Small-model function-calling pipeline that selects tools and generates typed parameters using constrained decoding instead of relying on free-form JSON generation. Built around Qwen3-0.6B with confidence-gated function selection and schema-driven output validation. Python · Qwen3-0.6B · Pydantic · Constrained Decoding
my_rag Retrieval-Augmented Generation pipeline focusing on semantic vector search, custom embeddings, and structured document context retrieval for LLMs. Active Research · Python · LLM Tooling · Vector Indexing

Engineering

Systems & Core: Low-level programming · Graph theory & pathfinding · Algorithm optimization · POSIX standards · Memory management AI & Web3: MCP · RAG · Web3/DeFi Infrastructure Stack & Tooling: C · C++ · Python · TypeScript · SQL · Bash · FastAPI · Convex · Docker · Linux · Git


Highlights & Education

  • 🏆 ETHGlobal Lisbon 2026 Winner — 3rd Place in The Graph's Best AI Tooling track for DeepTrace.
  • 🎓 42 Heilbronn — Software Engineering student focusing on core systems programming and software architecture.

Let's talk

Open to software engineering, systems, and Web3 roles. work@kapustazh.dev · kapustazh.dev · LinkedIn

Pinned Loading

  1. fly_in fly_in Public

    Fly-in is a graph-based drone routing simulator that uses capacity-aware pathfinding and real-time Pygame visualization.

    Python 3

  2. ikodo0/deeptrace ikodo0/deeptrace Public

    TypeScript 1

  3. SWOFART/OneShot SWOFART/OneShot Public

    OneShot protects an approved business obligation from duplicate committed settlements through retries, crashes, lost responses, queue redelivery, parallel workers, and multiple agent instances.

    TypeScript 4

  4. zelenskyy-codex-pet zelenskyy-codex-pet Public

    Unofficial fan-made Zelenskyy pet for the Codex desktop app, with full v2 animations and one-command installers.

    Python 4