Why we are hiring this role
On our AI World Tour (Charlotte, Vienna, Munich, Shanghai, Tokyo, Bangkok) one question came up at every stop: if agents write code this fast, who reviews all of it?
Our honest answer: for many low-risk changes we no longer read the code. We watch a short video of the new feature and press approve. Features ship behind feature switches, and users are the last check.
That only works if someone still looks properly. If you only ever press approve, you stop really looking. Taste is what stays human, and we want a Product Manager whose taste is our quality gate.
What we learned about taste
- Agents are fast, but not good at making something delightful. A human stays in the loop.
- You do not need to code. You need to say precisely what feels wrong and why, so an agent can fix it.
- Decide first, then ask the agent. Judgement gets weaker when you only approve. We wrote about it in Keep your judgement sharp.
- Quality gates for AI work (front end, back end, competitor comparison, AI behaviour) need someone who knows where the line is.
What to expect
- Own the product experience end to end: you define what good looks like and say no to what is merely fine.
- Review agent-built features by video, in feature-switch previews and from user feedback, and decide: ship, change or kill.
- Turn your reaction ("this feels off") into a clear brief that an agent and the team can act on.
- Work in our shared agent session with consultants, developers and designers.
- Set the quality gates: which automated checks run before a human looks, and what only a human can judge.
How we work: FAAFO
We work by FAAFO, from Gene Kim and Steve Yegge's book: Fast, Ambitious, Autonomous, Fun, Optionality. When agents carry refactoring, bug fixes and legacy, teams get time back for the ambitious things we kept pushing to next quarter. You get that room, and the expectation to use it.
What you bring
- Visible proof of taste: products, designs, writing or decisions you can show and defend
- The ability to explain in plain words why something is good or bad
- Product management experience. Technical depth is not required.
- Curiosity about working with AI agents every day
- Good English. German is a plus.
Your taste example
Bring one product, feature or page you think is great and one you think is broken, and tell us why, live. No code needed. Please do not share confidential material.
What we offer
- Unlimited tokens for Claude, OpenAI and the latest AI tools
- A team that works with agents every day and talks openly about what works and what does not
- A seat at the events where the field is being shaped
- Professional exchange in the XALT AI Circle and as an Atlassian Platinum Partner
Where you will meet us
- XALT AI World Tour 2026: six cities in seven weeks
- Enterprise AI Summit, Charlotte, 7-8 October: conference and our recap
- TEDAI Vienna, 28-30 October
- XALT AI Circle #2 with Peter Steinberger, Munich, 2 November
- AI Engineer Shanghai, 5-6 November
- Vibe Coding & Tech Networking, Tokyo, 15 November
- ProductTank Bangkok, 18 November
- Next: Enterprise AI Summit Portland, 5-6 May
Your application
Send your CV and your taste example to careers@xalt.de. The conversation also leaves room for your questions, and you can send them in advance to the same address.