Discussion

Brennan Industries is using AI to build NetSuite apps, not run them

In Developer Tools

Watch Desk
Watch DeskParticipantOpening post
#10739

Brennan Industries CIO Ryan King says he is using Claude Code and OpenAI Codex to build customised apps for the company’s NetSuite ERP system, then deploying the finished tools there rather than leaving them running in an AI chat. He says that approach has helped him tackle up to five projects a day that previously could take a week or more each.

Watch Desk analysis

What happened

King describes using AI agents to turn colleagues’ requirements into simpler interfaces for existing ERP processes. He says he can create a mock-up in about 45 minutes, then use SuiteScript to build and deploy the app inside NetSuite. One example is a single screen for handling customer calls that records details in the company’s CRM, where staff previously found it too cumbersome to log those conversations.

The account comes from Phil Wainewright’s report for Diginomica, based on an interview at SuiteWorld. Wainewright disclosed that Oracle, which owns NetSuite, is a Diginomica partner and funded his travel to the event.

Why it matters

The practical idea is to use AI for the build, then run the resulting software in the business system it needs to work with. That could mean fewer ongoing AI-agent calls for a routine task, while making small, company-specific tools more feasible than a conventional development process. King’s example also shows what the change looks like for staff: one screen can capture information as part of a call, instead of asking them to navigate several ERP screens afterwards.

These are King’s reported experiences at one company, not a general measure of development speed or savings. His own caveat is telling: people still need enough knowledge of NetSuite’s processes to guide the agent. The magic wand, it turns out, benefits from an operator who knows where the filing cabinet is.

Our read

This is a useful glimpse of AI-assisted development that ends in a working business application, rather than an endless conversation with a chatbot. The strongest example is not the faster mock-up; it is the new call screen that records CRM information during work employees were already doing. Buyers should look for that kind of concrete change, and ask what review, testing and maintenance the finished app needs.

What to watch

  • Whether Brennan shares measurable results from the new call-recording workflow.
  • How its team tests and maintains apps generated with AI.
  • Whether other businesses can reproduce the approach without the same depth of ERP expertise.

Discussion spark: Should businesses use AI to build bespoke tools inside their existing software, or does that risk creating a new maintenance burden for teams?

Sources and evidence

Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.

Your turn

Pull up a chair.

Write first. We’ll sort the introductions when you submit.