Senior Software Engineer (10+ yrs) — AI Tooling (MCP) · Flutter/Mobile · Rust/Solana
Currently exploring remote Lead/Staff roles in AI infrastructure, developer tooling, or mobile platforms at scale. GMT+4 (Tbilisi) — strong EU overlap, 4–6h overlap with US East.
- Cut AI translation costs by 99% — replaced $1,000/mo Crowdin with a $10/mo LLM pipeline (20+ languages, full app in 6 minutes)
- Shipped AI features to 20M+ users — chatbot handling 30K+ daily conversations with proactive suggestions
- Creator of mcp-rubber-duck — production MCP server for multi-LLM orchestration with tool isolation and guardrails
- Led teams of 3–15 engineers across crypto, healthcare, and enterprise B2B — fully remote
- Shipped Ever Wallet (100K+ downloads) with a Flutter-Rust bridge eliminating 100% of hand-written interop
Multi-LLM orchestration bridge — connected LLMs can use MCP tools directly without polluting the host context. Policy guardrails, usage tracking, multi-provider routing.
AI system prompt optimization — automated evaluation and iteration pipeline.
Flutter packages on Pub.dev
the_logger · the_storage · the_speech_to_text_button — logging, AES-256 encrypted storage, speech-to-text. Top pub.dev quality scores, used in production apps.
| Domain | Technologies |
|---|---|
| AI & DevTools | MCP, multi-LLM orchestration, RAG, prompt engineering, guardrails |
| Mobile | Flutter (7+ years), React Native, Kotlin, Swift |
| Systems | Rust, TypeScript/Node.js, Serverpod, PostgreSQL, Docker |
| Blockchain | Solana, smart contracts, Everscale |
| Infra | GitHub Actions, CI/CD automation |
MCP Rubber Duck series on dev.to: Fowler's GenAI Patterns Are Missing the Orchestration Layer · Stop Paying Twice for AI · Give Your Ducks a Fast Lane · Use a Rubber Duck Instead · Stop Copy-Pasting Between AI Tabs
Deep dives on Medium: When One AI Brain Isn't Enough · Full-Stack Solana Development · more...
Remote Senior/Lead/Staff roles in:
- AI developer tooling — MCP, agent infrastructure, model orchestration
- Mobile + AI — Flutter, cross-platform, LLM-powered features at scale
- Product engineering — high-ownership teams shipping real products





