Everyone says they work with AI
Appunite builds software for clients, and wants to be the firm businesses come to for AI. That means everyone has to be an expert at it, each in their own field. But there was no shared definition of what that means, so no benchmark and no shared way of working either.
First: what AI-ready means at Appunite
So the definition was the first thing we delivered together, in their own words and per kind of role.
- Engineering setting up your own AI environment the Appunite way, refactoring and reviewing code with AI, up to agentic engineering.
- Business using AI tools fluently every day, automating repeat work, testing ideas and creating documents.
What we measure
- Knowledge and skill
- Do you know what AI can do, and how to use it well?
- Use
- How do you really work with it, every day?
- Mindset and trust
- Do you dare to use it, and know when to check?
Everyone gets the same three dimensions, at the depth of the role: Core for business, about using AI and working with it. Build for engineering, about building with AI. Every score says what was measured and why, and the same answer always scores the same. So people compare, across roles and over time.
In hiring
When someone reaches Hired in Recruitee, the assessment goes out by itself, as a conversation, like their own interviews. The result is in Notion a week before the start date, with the reasoning, so onboarding is ready. The person gets their own result too.
Once the signal holds, the assessment moves forward, to right after the first phone screen. Then the recruiter has it for the interview. People keep deciding: nobody is rejected automatically.
In talent development
The same assessment comes back: after onboarding, before a review, and in between whenever a leader wants it. So AI skill becomes a line per person, and you see what onboarding achieved.
What Appunite can do with it
- Better matching Who works on which project, for which client, based on what someone can really do.
- More targeted training Onboarding starts where someone stands. And you see who needs what, and whether a training makes the difference.
- A promise to clients Once the readings hold, a bar you can promise your clients.
- Questions for your own data Everything is in Notion, and can be queried via API and MCP, with Claude for example.
The decision stays with Appunite. We run the assessment, the reminders and the results. A threshold only comes once the readings show it holds, and Appunite sets it.
