Harness engineering for AI agents
Design the system around the model.
Two lifecycle tracks for building reliable agents: product harnesses for quant research workflows, and coding harnesses for SaaS repo work.
Product Harness
Build a harness around an internal Quant Analyst AI Agent that produces research, backtests, advisory memos, review packets, and clean handoffs.
Starts with: Define the Work Surface 10 lectures Coding Agent Harness
Build a harness around a SaaS repo coding agent that handles issues, code changes, tests, CI, reviews, migrations, and handoffs.
Starts with: Define the Work Surface 10 lectures