Product & Usability Research Intern for T2D3 OS

The person who watches real users and tells us the truth.

The Research Intern owns the outside-in view of the product. AI can generate infinite UI and infinite copy; it cannot watch a real person get confused, hesitate, and quietly give up. You can. You run usability sessions, find the friction, and turn what you see into evidence the team acts on. The founder builds the product with Claude Code, the product lead shapes what gets built — you're the one who brings them the ground truth about how it actually lands. You come in curious, not finished.

This starts as an unpaid, semester-length internship built around your class schedule — the pay up front is real product experience, mentorship, and findings that change what gets built. It's built to grow: strong interns convert to a paid part-time seat, and become eligible for equity as the team grows. CS, HCI, design, and psychology-curious students all fit. Most internships won't look at you this early. We will, because we hire for trajectory, not résumés. Come in with curiosity, care, and a good ear, and we'll meet you at the start of the ladder below and grow together — teaching, pairing, and backing you the whole way.

What you'd do

• Run moderated usability sessions on onboarding and the core flows, and take careful, honest notes
• Turn what you observe into prioritized, evidence-backed findings — not opinions, not vibes
• Read basic product signal (funnel drop-off, Sentry web-vitals) to back a finding with a number
• Partner with the engineering intern to turn "this was confusing" into a reproducible issue
• Help decide what we test next, and close the loop by checking whether a fix actually worked

How it's different from a traditional seat

You're not filling out survey busywork that no one reads. Your findings go straight to the founder and product lead and change what gets built next. AI can draft the screens; you're the reason the right screens get drafted. What you're really building is judgment — the ability to separate what a user says from what a user does.

Where you might come from

First- and second-year students in CS, HCI, design, psychology, or anything adjacent — plus self-taught tinkerers who've always been the friend people hand their phone to and say "why won't this work?". A genuine curiosity about why people get stuck is worth more to us than a long résumé. You don't need research experience; you need to want to understand people.

You might be a fit if…

• You can say why a flow is confusing and what specifically you'd change
• You'd rather watch one real person use the product than guess with ten opinions
• You write findings other people can actually act on
• You're honest — you report the awkward truth even when it's about work someone loves
• You already use AI tools and are clear-eyed about where they help and where they mislead

The stack you'll work in

The live T2D3 OS product itself, plus the tools around it: session recording and note-taking, funnel and drop-off signal, Sentry for real front-end performance data, and Claude Code and MCP agents as part of the daily working environment. You'll learn it by doing the work next to the people who own the product.

How you grow as a Scribe

Investigative journalist, researcher, and storyteller. AI can rearrange words it has already seen; the Scribe uncovers the primary truth — the unspoken customer phrase, the real reason a flow fails — that AI cannot invent because it exists nowhere else yet. Interns join at the ground floor and grow from there.

Drafter, Level 1 of 4

Produces clean, clear findings and notes with AI assistance.

• Captures sessions accurately and writes them up clearly and without spin
• Every finding has a stated "what happened" and "why it matters"
• Reliable turnaround, session after session

Level up: Stop just recording sessions — go find the friction no one has named yet, and own one flow's research end-to-end.

Researcher-Interviewer, Level 2 of 4

Extracts signal from real users — sessions, interviews, primary observation.

• Runs sessions that surface an unspoken truth the team hadn't seen
• Captures the exact words and moments where users get stuck
• Backs findings with real signal and lets AI handle the secondary research

Level up: Turn one-off findings into a system — build a living map of where the product creates friction and set the bar others research to.

Narrator, Level 3 of 4

Owns the point of view on how the product actually lands; sets the research standard.

• Maintains the living canon — personas, journey notes, the friction map
• Produces insight the team can't get anywhere else because it rests on primary observation
• Trusted to say what the product gets wrong, and heard when they do

Level up: Stop generating findings and start generating understanding — define how the whole team thinks about the user.

Voice, Level 4 of 4

The org's understanding of its users is theirs; the research itself changes the roadmap.

• The team's picture of who it's for and what they need is one they authored
• Answers user questions no one else has answered well
• Insight alone moves what gets built

Level up: Ceiling rung; the move is outward — mentor the next researcher and put the standard into the system so it outlives you.

A friendly note on the journey: the real unlock isn't running more sessions than anyone else — it's the judgment to tell what a user says from what a user does, and to know which friction actually matters. That's a muscle we help you build.

How to apply

Tell us about a time you watched someone use something — an app, a form, a game, a tool — and spotted why it was confusing. What did you notice, what would you change, and how would you know the change worked? If you've ever done this with something you built yourself, even better.

We read for judgment, not buzzwords, and we reply to everyone who writes something real.