Technical Founder
nodes.garden / 2025-2026
Built a Web3/AI infrastructure platform from zero to 18,000+ users, 25,000+ deployed nodes, and six-figure revenue. Owned architecture, product delivery, and a distributed team of ten.
Senior Backend Engineer with fullstack range
Ilya Lebedev
I build and scale production SaaS, from backend architecture and data pipelines to frontend delivery and infrastructure.
8+ years across Ruby on Rails, PostgreSQL, background processing, APIs, infrastructure, and AI integrations. Former technical founder, comfortable owning ambiguous product work end to end.
Open to senior backend and fullstack engineering roles
Remote | CET timezone
8+ years
Production software
18K+ users
Platform scaled
25K+ nodes
Infrastructure deployed
Techstars
Web3 '25
Experience
Recent roles
Fullstack Ruby Engineer
Reinvented Hospitality / 2024-2025
Built and launched a Rails and Hotwire point-of-sale system in under three months, including operational workflows, tax reports, and exports.
Backend Ruby Engineer
Xometry / 2022-2024
Shipped regional marketplace launches, asynchronous intercompany systems, Salesforce pipelines, and an AI-assisted HS code search that reduced logistics search time by over 90%.
Flagship work
nodes.garden
LIVEMultichain Node-as-a-Service platform built through the full company lifecycle: product discovery, Rails architecture, infrastructure automation, smart contracts, hiring, and growth.
Reached 18,000+ users and 25,000+ deployed nodes, secured an Arbitrum ecosystem grant, and joined Techstars Web3 '25.
Rails / PostgreSQL / GoodJob / Solidity / Foundry / Docker / GCP / Fly.io
Recent technical work
Current hands-on builds
Smaller projects used to explore production patterns in Rails, agentic systems, RAG, and evaluation.
SupportOS
LIVEAgentic customer-support system for startup portfolios. Rails plus Hotwire, deployed with Kamal.
Ruby on Rails, Hotwire, SQLite, RubyLLM, OpenAI API, Kamal
slack-llm-bot
OSSProduction-shaped Slack bot on Vercel. OpenAI tool-loop, threaded memory via the Slack API, optional pgvector recall, and an offline eval harness (deterministic oracle plus LLM judge) so the LLM behaviour is tested like real software.
TypeScript, Vercel, OpenAI, RAG, pgvector
Core stack
Backend
Ruby on Rails, PostgreSQL, Sidekiq, GoodJob, REST, GraphQL
Product and frontend
TypeScript, React, Hotwire, Stimulus, Tailwind CSS
Infrastructure and AI
Docker, AWS, GCP, Fly.io, OpenAI API, RAG, LLM evaluation
Beyond work
Builder on the move
Currently in Almaty after visiting 30 countries. Former ETH miner with a 250+ GPU operation, PADI Advanced diver, freediver, FPV pilot, and board-game regular.
See travel map
Looking for a senior engineer who can own the whole problem?
I am open to full-time remote roles, strong product teams, and selective contract or technical co-founder conversations.