Blog
Field Notes
Practical notes on building AI agents, automation and AI-powered products for real businesses.
Securing Trust Between Agents in Procurement Workflows
Splitting a purchase across research, analysis, and execution agents makes the work easier to manage. It also creates more places where untrusted information can look like authority.
Securing Customer Support Agents That Can Issue Refunds
A support chatbot can explain a refund policy. A support agent can apply it, update a ticket, and move money. That changes what a security failure can cost.
AI Inside Your SaaS: Build Product Loops, Not Side Projects
How SaaS teams should put AI into the product itself — workflow by workflow — so customers feel leverage instead of another chatbot tab.
When Payment Agents Meet Adversaries: Why Verification Beats Vibes
Lessons from PayMAS Phase 0 — a synthetic research sandbox for measuring whether AI payment agents still honour user intent under prompt injection, data tampering, and tool poisoning.
Shipping Production AI Agents Without the Demo Hangover
A practical engineering view of taking agents from prompt playgrounds into systems that hold up under real users, messy data, and on-call reality.
START A PROJECT
Have something ambitious in mind?
We'll help you identify opportunities to improve your business with AI.