One MCP. The entire ERP.
Try Lensing with Claude, ChatGPT, or WorkBuddy


52-Min AI Build Session

Extend Your ERP Yourself: How Finance Teams Build Their Own AI Agents

Accounting teams can build their own tools now. The hard part isn’t building. It’s knowing what’s worth keeping, what needs controls, and what an auditor will ask about.

Sam Yen and Spencer Linder start with why finance teams end up extending their ERP at all, then separate AI features, agents and extensions to the core. They finish by walking through an application and an agent built on a live ERP, with every output traced back to source.

What You'll See

Every ERP is built on best practice, so it fits most of your business and never all of it. The gap gets filled with spreadsheets, bolt-ons and customizations.

Each customization makes the system harder to change. AI changes how that gap gets closed, but only if you can tell an AI feature from an agent from a change to the core.

In one hour: why extend your ERP at all, how AI features, agents and core extensions differ, the lifecycle from design to support, and an application and an agent built on a live ERP. This is the third session in our series, and the first built inside a real system of record.

Built for finance and accounting teams, not developers. No architecture diagrams.

  • Features, agents, or the core

    How to tell an AI feature, an agent and an extension to the core apart, and when each one fits.

  • Prototype discernment

    Where the line sits between a prototype worth hardening and one that shouldn’t go near the books.

  • An app and an agent, worked through

    An application and an agent built on a live ERP, from plain-English spec to a tested, traceable output.

Your Hosts


“Lensing's Founding Product Manager, who spent five years running advanced ASC 606 and NetSuite Advanced Revenue Management implementations for public and private companies.”

“Sam Yen — GM of AI at Lensing and formerly Chief Design Officer at SAP.”