McClaw — AI Tool
Find which local LLMs actually run on your hardware for AI agent setups. Hardware-aware model recommendations with real benchmarks across 340+ models.
Dmitry (Dee) Kargaev — AI Engineer & Product Designer · Los Angeles
I started freelancing in design at 14 in Irkutsk, Russia. By 20 I was leading a team at Spacecode, a digital agency shipping apps for Kaspersky, OTPBank, and 50+ other clients.
Then I spent five years as lead product designer at VALK (now DatAI Network) — a fintech platform processing over $4 billion in transactions, used by 70+ banks across 15 countries. Five industry awards. Coverage in CNN, Forbes, Yahoo Finance. I designed the product people at those banks actually used every day — compliance workflows, real-time risk dashboards, cross-border transaction interfaces.
Now I build AI systems. Not pitch decks about AI — actual systems running in production. Multi-agent architectures doing real work. I run dee.agency, a solo studio where I take founders from idea to working product with design, code, and AI integration. Fixed price, no meetings about meetings.
I write about what actually works in AI engineering at my blog — the stuff nobody puts in the pitch deck. Distribution problems, debugging agents, why most AI demos are cope.
I've worked from six countries, shipped products used by millions of people, and I still do all the work myself. No team, no layers, no project managers. You talk to me, I build it, you get a live URL.
BS in Computer & Information Systems Security — Moscow State Linguistic University
Find which local LLMs actually run on your hardware for AI agent setups. Hardware-aware model recommendations with real benchmarks across 340+ models.
Curated directory of tools, plugins, and apps built for the OpenClaw ecosystem. Browse, submit, and discover with X auth and community voting.
LinkedIn for AI agents. Discover, compare, and hire sub-agents for OpenClaw and other orchestration frameworks. Clerk auth + Convex backend.
Intelligent sub-agent spawning for OpenClaw. Auto-selects the right model for each task based on benchmarks, cost, and capability. 340+ models, 5 data sources.
Give your AI agent access to any API in ~1,000 tokens. Two tools replace hundreds of definitions. 56 APIs ready out of the box.
Astro static blog with an automated content pipeline. Scout → rank → draft → SEO review → publish. Fully autonomous content generation for OpenClaw ecosystem.
GPU inference in one command. Picks a cheap GPU on Vast.ai, loads any GGUF model, gives you an OpenAI-compatible endpoint. Auto-kills on idle.
React, Next.js, TypeScript, Tailwind, shadcn/ui, Vue, Svelte, Astro, Vite, Remix, Nuxt, Turborepo
Product Design, Design Systems, Figma, Brutalism, Interaction Design, Information Architecture, Design Tokens, Component Libraries, Accessibility, Responsive Systems
Multi-Agent Orchestration, RAG Pipelines, LLM Tool Calling, Prompt Engineering, Embeddings, Vector Search, Eval Frameworks, Agentic Workflows, Function Calling, Context Window Optimization, Fine-Tuning, Semantic Search, Chain of Thought, Knowledge Graphs
LangChain, LangGraph, CrewAI, n8n, OpenClaw, Cursor, Claude Code, Codex, OpenRouter, MCP Servers, Claude, GPT, Gemini, Grok, Ollama, Llama, Mistral, Qwen
Fintech, DeFi Protocols, Tokenomics, Solidity, Institutional Trading UX, Custody Systems, Compliance Interfaces, Web3, Smart Contracts, On-Chain Analytics
TypeScript, Python, Swift, SwiftUI, Solidity, SQL, Node.js, GraphQL, WebSockets, Edge Functions
Supabase, Convex, Postgres, Redis, Pinecone, Drizzle, Prisma, Vector Databases
Framer Motion, Three.js, Matter.js, WebGL, GSAP, Spline, React Three Fiber, Shader Programming, Physics Simulations, Procedural Generation
Docker, Vercel, Railway, Hetzner, Caddy, GitHub Actions, Cloudflare, Tailscale, Syncthing, Self-Hosting, Telegram Bots, Discord Bots, Chrome Extensions