Job-Hunting in Public, Without Needing a Job

Bear is a professor with a job, running a job search in public. Why? Two reasons, one project.

He doesn't need a job — he wants consulting work, and he's short on public examples to show for it. So he's making them, in his classes and on this playlist, where the real problems land on camera. He's an Associate Teaching Professor in AI and machine learning at Northeastern, and his folder holds a saved job board, his CV turned into structured data, and a gap table — the shape of a real search.

Two reasons, one project. First, a pitch: consulting work, for himself or his company, Bear Brown and Company. Second, a lesson: the same job-search tool is built live in three of his classes — Computational Skepticism, Branding and AI, and Prompt Engineering — with identical code and only the question changing. Skepticism asks how you know the output is right. Branding asks whose problem it solves. Prompt Engineering asks what the recipe is and which labor belongs to Claude. Computational Skepticism holds the master copy, synced out to the others, because the hard part isn't fetching jobs — it's knowing what the output is worth.

The tool is called Lectern. Its collector reads eighteen job boards across three applicant-tracking systems, keeping only postings about teaching.

The pitch itself: companies with AI products want to sell to universities, but most don't understand how a university works — its politics, its needs around AI. Bear proposes a connector: someone who teaches AI, sits on the committees where a university decides its AI direction, and actually uses the tools. His CV, turned into structured facts and quoted directly rather than from memory, backs part of that — teaching, course design, a dean-commissioned course, a Coursera series, a nonprofit fellowship. A gap analysis against one real posting — Figma's Designer Advocate, Partnerships — found strong teaching and program-building, but no mention of Figma, design systems, certification programs, or conference talks. His own framing: what's missing are concrete public examples, not credentials.

A scan of Figma's own job board that day found two Designer Advocate roles, one stated as full-time, and no posting offering part-time, contract, or university terms. The resulting recipe, labeled a judgment rather than a record: network, don't apply.

That's where this playlist comes in — if the gap is examples, build the examples. Figma for Educational AI is planned as a book, a set of workshops, and this playlist, all on one Figma for Education account. And the films keep the friction in: a revoked login, a leftover MCP server shadowing a plugin's Connect button, a FigJam board that flattened its own labels. That's not the series failing — that's the series. A tutorial that hides its problems teaches nothing about the ones you'll actually hit.

Your Turn: Paste into Claude, with a real posting and your CV: "Here is a job posting and my CV. Quote the posting's top five requirements. For each one, quote the line in my CV that meets it, or write: not in my CV. Then name one public example I could build to close the biggest gap." Then check two things yourself: is every requirement quoted from the posting, and is every match quoted from your CV?

Sources: the job board scan, gap analysis, and CV facts are public in Bear's course folders — Prompt Engineering example folder and Computational Skepticism master tool repo (github.com/nikbearbrown). Claims Bear makes about himself that aren't in his CV facts file are stated as his own. An independent series, not affiliated with or endorsed by Figma.

0:00 A professor job-hunting in public 0:38 One project, two purposes 1:04 Three classes, one tool 1:27 Lectern, the job-search tool 1:40 The pitch: Bear's argument 2:06 The evidence, from his CV 3:11 The gap analysis 3:47 What the job board said 4:12 Network, don't apply 4:22 Where this playlist comes in 4:34 The friction stays in 5:05 What's been made so far 5:36 Recap 5:51 Your turn

Series: Figma for Educational AI. @NikBearBrown

job search in public, AI consulting pitch, CV gap analysis Claude, job posting requirements match, Figma Designer Advocate, university AI consulting, teaching examples portfolio, Claude Code job search tool, Lectern job board collector, Computational Skepticism AI, quoted CV analysis, network don't apply, Figma for Educational AI series, building public examples, AI in higher education

What this video covers

• 0:00 — A professor job-hunting in public 0:38 One project, two purposes 1:04 Three classes, one tool 1:27 Lectern, the job-search tool 1:40 The pitch: Bear's argument 2:06 The evidence, from his CV 3:11 The gap analysis 3:47 What the job board said 4:12 Network, don't apply 4:22 Where this playlist comes in 4:34 The friction stays in 5:05 What's been made so far 5:36 Recap 5:51 Your turn

Transcript

A lightly formatted transcript of the video, with timestamps so you can jump to any section on YouTube.

A professor job-hunting in public 0:38 One project, two purposes 1:04 Three classes, one tool 1:27 Lectern, the job-search tool 1:40 The pitch: Bear's argument 2:06 The evidence, from his CV 3:11 The gap analysis 3:47 What the job board said 4:12 Network, don't apply 4:22 Where this playlist comes in 4:34 The friction stays in 5:05 What's been made so far 5:36 Recap 5:51 Your turn

[0:00] Aloha, this is Liam in for Bear. Bear is a professor with a job and he's running a job search in public. This film is the context for the whole playlist, what he's doing, why, and what you're about to watch. A professor job hunting in public is building his proof. Bear doesn't need a job, he wants consulting work and he's short on examples to show. So, he's making them in his classes and on this playlist where you can watch the real problems land. Start with the odd part. Bear is an associate teaching professor in AI and machine learning at Northeastern. He says he doesn't need a job and yet his folder holds a saved job board, his CV turned into data, and a gap table. That's a job search. Why run it in public? Two reasons, one project. First, a pitch. Bear wants consulting work for himself or for his company, Bear Brown and Company.

Second, a lesson. His students watch the same project hit real problems and watch him work through them. The lesson first, the same job search tool is being built in three of Bear's classes, computational skepticism, branding and AI, and prompt engineering. The code is identical in all three, only the question changes. Skepticism asks, "How do we know the output is right?" Branding asks, "Whose problem does it solve?" Prompt engineering asks, "What's the recipe and which labor belongs to Claude?" Computational skepticism holds the master copy. Shared code is edited there first, then copy it out to the other two with one sync script. Why there? Because the hard part isn't fetching jobs, it's knowing what the output is worth. The tool is called Lectern. It's collector.py reads 18 job boards across three applicant tracking systems, 12 on one, five on another, one on the third. It keeps only the postings about teaching. Now the pitch, this is

Bear's argument and it's his to make. Companies with AI products want to sell to universities, but most of them don't understand how a university works, its politics or its needs, especially around AI. So, he proposes a connector, a maven, someone who teaches AI, sits on the committees where the university decides its AI direction, and actually uses the tools. Someone who can stand in that gap and translate in both directions. Is that Bear? He says he teaches five AI courses. He says he has run and still sits on advisory committees on the university's AI direction. Those are his statements. The committee work isn't in his CV file. What follows is his CV turned into structured facts he read and attested. Headline, associate teaching professor, AI and machine learning, Northeastern University. Every claim from here to the gap table is quoted from that file. An engineering course, ENGR 0201, Dean commissioned, 700 plus learners. The CV calls

It the entry point course for Northeastern's university-wide Entropic Cloud Enterprise Partnership. 25 plus AI-powered course assistants designed and deployed across Northeastern courses and programs. And the algorithms course, INFO 6205, co-developed and filmed as a full Coursera course, 13 modules. Outside the classroom, he founded Humanitarian's AI with 150 plus fellows across 16 projects. And Bear Brown and Company builds AI tools for musicians and for educational applications. Then he ran a gap analysis against one real posting, Figma's designer advocate partnerships. On the left, what the posting asks for quoted. On the right, what the CV file says quoted. No memory, no adjectives. The finding, the CV shows strong teaching and program building, but never mentions Figma, design systems, certification programs, or conference talks. A word search of the file finds none of them. Bear's own framing is blunter. What he's missing are concrete examples, tutorials, workshops, educational materials made and out where people can see them. A

CV says he could. Examples would show he does. Then the job board itself. On September 23rd, Figma's public board listed 160 postings. A small script scanned everyone for advocate titles, education words, and flexible terms. The result, two designer advocate roles and the one that states its terms as full-time. Not one posting offered part-time, contract, temporary, seasonal, or freelance terms. None mentions universities. So, the recipe's next action, and it's labeled a judgment, not a record, network, don't apply. If contract or university work exists, it arrives through a conversation, not through a board. Which is where this playlist comes in. If the gap is examples, build the examples. Figma for educational AI is planned as a book, a set of workshops, and a playlist, all on Bear's Figma for education account. And the films keep the friction in. In one session, Claude Code's login was revoked, a 401 error, and a leftover Claude MCP ad server shadowed

The Figma plugin's own, so the connect button stayed gray. Figma reports his education team as student tier, and when Claude drew his project as a fig jam board, the board kept every box but flattened the labels. That's in the film, ring and all. That isn't the series' failing, that's the series. A tutorial that hides its problems teaches you nothing about the ones you'll hit. So far, five films, the playlist intro, Figma's MCP server, connecting Figma to Claude co-work, Claude Code connecting and drawing the project, and why a student should hand in the version history, not just the design. Behind them sits a test plan, 40 experiments on the education account. Each is written to end in a verdict, strong, middle, or weak. Watch for the verdicts, and watch for what broke. And watch the method, every value is labeled, a record, a judgment, or Bear's own input. The tools do the typing, a person makes

Every decision. Let's recap with Claude. Bear doesn't need a job, the search is a consulting pitch and a live lesson, one project in three classes. His argument, AI companies don't read universities, and a connector can his own gap examples. So, this playlist builds them friction included. Your turn. Paste this into Claude with a real posting in your CV. Here is a job posting in my CV. Quote the posting's top five requirements. For each one, quote the line in my CV that meets it or write not in my CV. Then, name one public example I could build to close the biggest gap. Why this prompt? Quotes on both sides mean nothing gets invented and every gap can be checked and it ends on one thing you can build. Not a list of regrets. Check each quote against the posting before you trust the table. Job hunting in public without needing a job at Nick Bear Brown.