What’s happening behind the scenes when AI gets verified?
In part 2 of Prove It, we explain how verification actually works. From formal logic and Cedar policies to Amazon Bedrock Guardrails and AgentCore, we unpack the technology and economics behind building trustworthy AI
Get started, scale up. For founders and everyone in the build.
Joined July 2010
- There's a new generation of founders in town. We got to chat with some of them at our @ycombinator Startup School afterparty. They talked about what they're building, what it means to be in San Francisco, and why this event brings people and ideas together. More compute. More
- What happens when your AI is wrong? “Probably correct” isn’t a compliance strategy. Prompt engineering, RAG, and manual QA can reduce risk, but they can't prove correctness. In our new startup series, we explore how Automated Reasoning checks in Amazon Bedrock help verify AI
- The AWS Impact Bootcamp Tokyo experience: Catrinel Hagivreta. “Today, there are so many opportunities for female founders,” says the Co-founder & CEO of startup MEDIjobs. Catrinel explains why she applied to the Bootcamp and how the two-day event will help unlock new
- Supply chain insurance just got a lot smarter. @ParsylHQ is changing how cargo risk is assessed, using AI on AWS to cut quote times from weeks to as little as two days. With sharper, data-driven risk insights, the startup is helping build more resilient global supply chains.







