Skip to content

AI in your organisation, doing real work from your own data

We put AI assistants and agents to work in your organisation: answering reporting questions, preparing management packs and drafting documents from your own data, inside the tools your team already uses. Every answer shows its source, and nothing goes live until it passes questions you already know the answers to. We start with a proof of value on work that matters to you, then take it further: commentary, alerts and first drafts.

Why iRonin: 19 years of building software for companies, and AI agents already doing real work in our own delivery, not in a demo.

What you can check

  • Our fixes are merged into Hindsight, the open-source agent-memory project with over 62,000 GitHub stars. #2611 → · #3144 →
  • 25 of our pull requests are merged into open-source projects run by others, 19 of them into AI-agent tools. See them on GitHub →
  • Rated 4.9 out of 5 on Clutch, from 33 reviews of our software and engineering work. Read the reviews on Clutch →
Who it is for
Finance, operations, legal, sales and any other team that wants AI working from its own data.

What you get

  • A working assistant where your team already works — Teams, Slack, SharePoint or Google Drive first; a web workspace for heavier jobs.
  • A glossary of your metrics and terms, and a written definition of what it answers from which data — confirmed by one of your people.
  • A golden set first: 20–30 questions you already know the answers to. Passing it is the test before rollout.
  • The source shown with every answer, and for data questions the query behind it.
  • A plan for the next step: from answering questions to doing work — commentary, alerts, first drafts.

How it works

  1. Plan

    The work, and the answer we will measure it against.

  2. Design

    Agents explore your data and draft the definitions; one of your people confirms them.

  3. Execute

    Built into the tools your team already uses, with a web workspace for the heavier jobs.

  4. Verify

    Proven against questions you already know the answers to, before rollout.

Built with

Claude, GPT, Gemini or open-weight models; your cloud; inside Microsoft, Google or open-source stacks. We do not sell software licences.

Under way

A proof of value for a finance team on Microsoft's data stack.

Questions

  • Do we need a data platform first?

    No. Agents explore the data you already have and draft the definitions; your team confirms them.

  • Does our data leave our environment?

    Not unless you want it to. It can run in your own cloud, with the model you choose.

  • How do we know the answers are right?

    Every answer shows its source, and accuracy is proven on questions you already know the answers to before anyone relies on it.

  • Can we see it work on our own material first?

    Yes. The work starts with a proof of value on one of your own tasks, measured against answers you already know.