A lab formarket intel
We build the tools we'd want to use ourselves — non-custodial financial intelligence and open-source infrastructure for the agent economy. Both products live as public MCP endpoints, plug directly into any AI client.
Two products,
one lab.
Both are live, both are public, both expose an MCP endpoint you can point an AI client at today.
Root Financial
Non-custodial trading intelligence for modern markets. Live data across DEXes, perps, options, equities, and prediction markets — surfaced as a public MCP endpoint that plugs directly into Claude, Codex, and the OpenAI API. Wallets sign transactions locally; we never touch keys.
Cult OS
Open-source infrastructure for repository work in the agent economy. Explicit contracts link issues to agent deliverables; verification, x402 USDC micropayments, and on-chain settlement receipts flow through a single MCP endpoint. Maintainers keep control — no auto-merge.
Plug both MCPs
into any AI client.
Root Financial and Cult OS both expose MCP endpoints. Point Claude Desktop, the Codex CLI, or your own OpenAI API code at the URLs below and the tools become native to your session.
1{2 "mcpServers": {3 "root-financial": {4 "url": "https://root.financial/mcp"5 },6 "cult-os": {7 "url": "https://cultos.dev/mcp/connect"8 }9 }10}
What actually
exists today.
Two operators, one lab.
Financial engineer. Since 2018, building production algorithmic-trading systems and the data + execution infra behind them.
Since 2018, I've built and managed production-grade algorithmic trading systems and the infrastructure required to operate them continuously and reliably. My experience spans quantitative research, automated execution and hedging, options and volatility strategies, large financial datasets, cloud infrastructure, and integrations across both centralized crypto exchanges and traditional brokerage APIs.
As AI and LLMs emerged, I was an early adopter of AI-assisted development and saw firsthand how dramatically it could increase engineering and research productivity. That experience also exposed a new bottleneck: models have become increasingly capable, but connecting them reliably to specialized data, tools, APIs, and production workflows remains fragmented.
Root Systems grew directly from that intersection. We are applying my background in financial engineering, data science, and production infrastructure to build standardized AI and agentic tooling that improves how developers, quantitative researchers, funds, and other technical teams access data, conduct research, build software, and deploy reliable workflows.
Talk to us.
We're raising for the next phase — deeper data pipes, agent orchestration, and the connective tissue between both products. If any of this rhymes with your thesis, get in touch.
