Self-taught engineer with a lab sciences background. I build production systems — autonomous research agents, trading infrastructure, and hardened local AI runtimes — and ship them.
A hardened Windows installer that packages an AI agent runtime for non-technical users. Wraps the agent in a sandboxed Linux environment with an egress firewall, encrypted API key storage via Windows Credential Manager, a loopback-only gateway, and a kill switch. 17 build iterations. Full Azure headless validation. Two products: ClawFactory ($149, multi-agent) and ClawAgent ($49, single agent). Under Frontier Automation Systems LLC.
A live subscription platform ($49/month) for macro investment intelligence. The core is a regime engine that classifies the macro environment across 7 pillars (Growth, Inflation, Liquidity, Rates, Credit, Dollar, Risk) into one of 5 states — Goldilocks, Reflation, Stagflation, Disinflation, or Transition — with confidence gates and persistence rules to prevent false signals. Delivers portfolio allocation guidance, stress indicators, and an AI research copilot. Built on React, Express, PostgreSQL, and Railway with live data from FRED, Polygon.io, CoinGecko, and Stripe.
A 21-agent autonomous research system running 24/7 on a local machine. Assigns itself research objectives across macro, crypto, geopolitics, and credit markets — searches the web, synthesizes findings, stores everything in a structured knowledge base, and self-corrects when stuck. Over 8,500 research objectives completed without human intervention. Operates on a weekly rhythm with dedicated regime classification every Monday. Feeds a publishing pipeline that delivers intelligence to Avital Research. Built in Python with PostgreSQL, Pinecone vector storage, and a React mission control dashboard.
Single-agent local AI runtime. Same security substrate as ClawFactory — WSL2, egress firewall, DPAPI — faster install.
Native Windows desktop chat interface for local AI agents. Conversation history, dark/light themes, streams from local gateway.
Autonomous strategy research and backtesting pipeline. Generates, tests, and ranks trading strategies across macro regimes. BTC, ETH, SOL, XRP.
How the pieces fit together. An autonomous intelligence engine does the mechanical research; a subscriber-facing platform distributes only validated output; a human stays in control at every publish.
Avital Research runs a macro research copilot grounded in the regime engine. Ask it something and it opens in the live app.
Opens the live copilot on Avital Research.
Study Director at Nelson Laboratories, Salt Lake City. Lab sciences background, self-taught in software. Building a fully autonomous AI organization on the side — research agents, trading infrastructure, publishing pipeline, and the tools to run all of it locally with real security controls.
A small shell. Type help, or try open avital.