End-of-semester with Anna, professor who endorses six projects in twenty minutes
Professor endorsement is not the moat. AI extraction from real artifacts is. But when professors do have time to weigh in, their endorsement adds a trust layer that makes a good signal even better. Here's how Anna does it without it becoming another commitment.
The thing we don't ask professors to do
Most career-service systems implicitly require professor involvement for credibility. The professor verifies, the system displays the verification, students benefit. The problem with that model is that professors are busy, and a system that depends on their consistent participation is fragile.
InTransparency doesn't work that way. The skill verification happens through AI extraction from the artifact itself — the code, the thesis, the design files. A student's profile is credible without a single professor touching it.
Professor endorsement exists, but it's an optional bonus signal. When a professor adds an endorsement, it adds a trust layer on top of an already-evidenced profile. When they don't, the profile still works.
This is Anna's day with that bonus layer.
End of semester, Anna's lab
Anna teaches Computer Engineering at a polytechnic. She runs a lab section on embedded systems for second-year students — about 35 students per cycle. End of semester means project demos: each team presents a small embedded project they've built over the term. Anna grades the demos as part of the course.
What Anna doesn't normally do is write LinkedIn recommendations or fill out career-service forms. She has 35 students this term, 40 the term before, and 45 the term before that. The math of "write a personalized recommendation for each" doesn't work.
The endorsement screen
After demos she opens InTransparency. The system shows her every project from her course that students have uploaded — 28 of the 35 have. Each one is annotated with the AI's extracted skill list and a link to the artifact.
For each project she has three options: skip, endorse, or endorse with comment. Skipping is fine and explicit. Endorsing is one click — it adds her name to the project as the supervising professor and stamps a trust signal that recruiters will see.
She decides to endorse six projects this cycle. They're the ones where the demo went well, the documentation was clean, and the student demonstrated more than the basic course objective. For two of them she adds a one-sentence comment. For four she just clicks "Endorse." Total time: about twenty minutes.
For the other 22 projects, she skips. That's fine. Those students don't get her endorsement, but they also don't lose anything they would have had — the AI-extracted skills are still on their profile, the artifact is still linked, and recruiters can still find them.
What recruiters see
When a recruiter looks at one of the six endorsed projects, the profile carries an extra small badge: "Endorsed by Prof. [Anna's name], embedded systems, 2026 spring cohort." If the recruiter clicks the badge, they see the one-sentence comment if there is one, and a verification that this professor taught this course this term.
That's it. No five-paragraph testimonial. No formal letter of reference. Just a clear, clickable signal: a real instructor, who taught the relevant course, has put their name to this work.
For the recruiter, this is useful in the same way a clean GitHub commit history is useful: it doesn't replace their evaluation, but it shifts the prior.
Why this design
The design is intentional. We considered making endorsement a heavier-weight, more formal artifact — full recommendation letters, structured evaluations, multi-question forms. We rejected it because every additional minute of professor commitment reduces participation, and reduced participation makes the signal noisier (the professors who endorse become a self-selected group).
A one-click endorsement that takes 20 seconds means more professors will use it. The signal is weaker per endorsement than a 500-word letter, but it's available across more students, so the population-level value is higher.
When professor endorsement isn't there
Most students on the platform do not have a professor endorsement. Their profiles still work. They are still discovered by recruiters. Their skill evidence is still extracted from real artifacts. A recruiter looking at a Marco profile without an endorsement reaches the same conclusion they would have reached with one — the artifact is the artifact.
The endorsement, when present, is a bonus. It's not the foundation. The foundation is the artifact, the AI extraction, and the platform's audit log.
That distinction matters. It means the platform doesn't depend on Anna's consistent participation, but it also means that when Anna does have twenty minutes at the end of a semester, those twenty minutes produce real value for six of her students — without becoming another permanent obligation on her calendar.
That's the platform from a professor's seat: a tool you use when you have time, not one that adds a new commitment to a teaching load that's already full.