Filed under Agents.
Notes, numbers, and hard-won lessons from helping small and medium-sized businesses actually put AI to work. No think-pieces, no vendor takes — just what the math said after we ran it.
3 Million Tasks Nobody Had to Do
C.H. Robinson's AI agents completed 3M+ shipping tasks. What happens when a 40-person logistics broker applies the same math to their own operation?
Why Process Beats Raw Capability: What Andrew Ng’s Agentic Workflows Mean for Your Operations
GPT-3.5 in an agentic loop scored 95.1% on coding benchmarks. GPT-4 zero-shot scored 67%. The cheaper model with a better process won. Here's what that means for your operations.