Insights on agentic AI & product engineering
Field notes from the systems we build and ship.
What US Policyholders Actually Trust AI to Do (And What They Don't)
Support for AI in insurance nearly doubled year over year, but consumer trust has clear limits. Here's where US policyholders welcome automation, and where insurers risk real backlash by pushing too far.
Why Insurers Are Ditching Monolithic Claims Platforms for Composable Architecture
The all-in-one claims platform promised to handle everything and often delivered a system too rigid to adapt. Here's why carriers and MGAs are moving to composable, API-connected claims automation instead.
AI Fraud Detection Isn't Just for Big Carriers Anymore
Enterprise-grade fraud detection used to require a dedicated data science team and a large carrier's budget. That's changed. Here's what mid-size insurers and MGAs can now deploy, and what it actually catches.
APRA's AI Letter Just Changed the Rules for Australian Insurers — Here's What It Actually Requires
APRA's April 2026 letter to industry put every regulated insurer on notice: AI governance is now a prudential risk, not an IT decision. Here's what the letter actually requires, in practical terms.
Building Insurance AI That Satisfies Regulators in the US and Australia
US and Australian insurance regulators are converging on the same demand: prove your AI is governed, not just functional. Here's what a compliance-ready AI build actually looks like in both markets.
From Chatbot to AI Agent: What's Actually Changed in Customer Support Automation
"Chatbot" and "AI agent" get used interchangeably, but they solve different problems. Here's the real difference, and why it matters for what your support automation can actually deliver.
What Enterprises Need in Place Before Scaling AI Agents Past Pilot
Security and governance are now the top factor enterprises weigh when evaluating agentic AI. Here's what actually needs to be in place before an AI agent gets real autonomy in your business.
AI Automation for Enterprise Workflows: Where to Actually Start
Enterprise teams are under pressure to automate everything with AI agents at once. Here's how to pick the right workflow to automate first, and avoid becoming one of the projects that gets shelved.
Your Company Probably Runs 12 AI Agents That Don't Talk to Each Other
The average enterprise now runs a dozen AI agents, but half of them operate in complete isolation. Here's why that's quietly limiting your automation, and how orchestration fixes it.