for faculty

Teech AI for faculty

We interviewed faculty at leading universities about where their week actually goes, and identified the pain points that come back every time. The teaching itself is fine. The work around it is not: repeating emails, exams marked one by one, and material that never quite fits the class in front of you.

Three building blocks. Pick one to see it working.

Non negotiable

Two conditions we were given

Both came up unprompted in every interview. Everything above is built around them, and you will see both in every demo.

The professor sends. Not the system.

One wrong email to the entire department costs more than the hour it saved. So nothing sends, posts, submits or grades on its own.

Drafts are prepared and queued. The professor reads, edits if needed, and approves. Approval is one click, and it stays a click a person makes.

The data stays where the university wants it.

The hesitation we heard was not about quality, it was about connecting a mailbox and student data to something outside the institution.

So: deployment inside the university's own cloud account, content never used for training, and each connection approved separately rather than all at once. A workflow can also run on files the professor hands over, with no live connection at all.

Budget

Where the money for this already sits

None of this needs a new budget line to start.

RouteWhat we foundFits
Institutional licencesThe universities we spoke to already pay for general AI assistants, and for specialist legal tools on the law side. Paid AI tooling is established, so the question is which workflow, not whether.Department wide rollout
Research fundsFaculty hold personal research funds, often in the region of 10,000 euro a year, and are reimbursed for subscriptions they buy themselves. No central agreement needed.Single professor pilot
ProcurementSeveral universities run an innovation procurement unit whose job is finding and testing solutions like this, which is the route to more than one department at a time.Multi department contract
Pilot

How we would start

Six weeks, one professor, two workflows, measured.

Week 0

Pick two workflows.

Usually admin mail and exam marking, because both repeat weekly and both have a clean before and after.

Weeks 1–2

Two weeks to build.

Built on that professor's real cases, their templates, their phrasing, inside the university's own environment.

Weeks 3–6

One month running.

The professor reviews and approves every output. We count hours returned, drafts accepted without edits, and anything the review caught.

Week 6

Then decide.

Extend to the rest of the department's admin portfolios, or stop. The pilot is sized so stopping is cheap.