Agentic Options Trading Platform - Part 1

In this episode of The Build, Cameron Rohn and Tom Spencer discuss building an agentic options trading platform and the engineering choices that enable it. They begin by grounding the project in practical constraints: options market data, broker integrations, and the trade-off between latency and reliability when agents interact with live APIs. The conversation then shifts to tooling and workflows, highlighting Langsmith for agent orchestration and observability, Vercel for deployment, Supabase for rapid database and auth, and MCP tools for prototyping integrations. They explore memory systems and AI agent design, weighing session-level versus long-term memory and how that shapes stateful trading strategies. They explore technical architecture decisions around API integration, real-time data ingestion, and modular services that separate a UI layer from broker-facing adapters. They explore building in public as a growth and product-feedback strategy, discussing transparency, monetization experiments, and community-driven open source components. Later, they address developer ergonomics: local-first workflows, CI/CD, telemetry, and cost-conscious scaling. The episode closes with entrepreneurial insights on iteration velocity and risk management. The key takeaway for developers and entrepreneurs is forward-looking: favor modular, observable architectures and leverage modern tools to ship agentic systems quickly while learning in public.