I'm Conor Svensson — a technical founder in London. I spent a decade building trading and regulatory reporting platforms for investment banks, and the last decade with my company Web3 Labs building developer infrastructure for blockchains. These days I write here about how AI is changing the way I build things.
-
You can't manage AI risk you haven't measured
Regulation and standards are emerging for building applications that use LLMs. Combined with the right approach to evaluation, they let you quantify your risk exposure.
-
Why you need an operating model for agentic engineering
Technology teams used to standardise on frameworks and build systems. With much of this work delegated to agents, you need to standardise on how you use them. This is where an operating model is crucial.
-
Building an evals framework for audits
The document-heavy nature of audits makes them well suited to using LLMs. But in regulated industries you need the right guardrails in place before you unleash them.