Bear Brown

Talent

Bear Brown & Co.

A professor sees the best. Hire them away, please.

The year, and the gap in it

I am a professor. I teach thousands of students — AI, machine learning, data science — in the College of Engineering at Northeastern. Over a few semesters you stop guessing: you know which ones are actually good, as opposed to which ones interview well. Those are different lists.

Here is what nobody tells the good ones. Landing a job at a big tech company takes roughly a year, even for the best of them. Application, recruiter screen, take-home, four rounds, team match, offer, a start date another three months out — and that is the path when it goes well. It is not a verdict on the candidate. It is how the pipeline runs.

So a genuinely strong AI engineer spends a year underemployed, watching the edge they just spent two years sharpening go dull. Meanwhile there are people with something real to build — a tool, a dashboard, an agent, a working version of an idea — who cannot find anyone to build it at a price that makes sense.

Those two problems are each other's answer. Temporary, paid, real project work for AI and STEM graduates during the year they are job hunting; working software for the people who need it built. I am the connector. I know which students to send, I supervise the work, and if you want to keep the person at the end of it, you keep them.

Temporary is the feature, not the compromise. The work is scoped to the months a graduate is job hunting. If it turns permanent because you hired them, the arrangement did exactly what it was built to do.

What we do

We build AI-era web apps — interactive tools, dashboards, assessments, wrappers around your models and data — using open-source tools and recent graduates from top AI programs.

The code is real work, delivered fast and cheap. But the bigger thing we offer is a structured way to audition top AI talent on a real project. If you want to hire the person who built your app, we'll introduce you. No placement fee. No non-solicit. That's the point.

Two lanes. No middle.

Open Lane

$35/hr

For the 95% of what most clients need: wrapping your vision in open-source tools so people can actually use it.

Recent grad developers, supervised by Prof. Nik Bear Brown (Northeastern University, College of Engineering)
$35/hr, billed weekly
Open source only — MIT, Apache, BSD dependencies
No NDAs. If your project needs one, it belongs in the Bespoke Lane.
Client never shares proprietary IP with us. We build the wrapper; your secret sauce stays yours.
Client owns the code and can do whatever they want with it.
No warranty period. Bugs get fixed at the hourly rate. Retainers available.
Either side can stop at a week boundary. No long-term commitment.
Hire the grads. If you want to bring the developer in-house, we'll introduce you at no cost. If it happens mid-project, we cover the shadowing cost for a clean handoff.

Bespoke Lane

$200+/hr or equity

For the 5% that genuinely needs it.

Nik directly, possibly with vetted senior collaborators
$200+/hr, equity, or both — priced per engagement
NDAs, exclusivity, direct IP handling — all on the table
Meaningful engagement size. Not a weekend gig.

Why this works

For clients: Essential work done — and a possible first employee. You get a working app at a rate no dev shop can match, built on infrastructure you can maintain yourself, by someone you can hire if they're a fit. The structure of the engagement is itself the interview.
For grads: Paid work during the year the big-tech search eats, real project experience instead of a résumé gap, a professional reference, and a direct path to a job with a client who has already watched them work.
The moat: The economics only work because Nik teaches thousands of students and knows, first-hand, which ones are strong. No dev shop can replicate that vetting, and no résumé screen can either. That's the moat, stated plainly.

What we build

  • Interactive assessment tools (sliders, spider charts, scoring dashboards)
  • Data viz and executive dashboards
  • LLM-wrapper applications over your existing models, data, or workflows
  • Agentic systems — multi-step automations, tool-using agents, workflow orchestration
  • Research tools — literature review, data gathering, analysis pipelines
  • Fine-tuned models on your open data
  • Internal tools, client-facing portals, prototypes
  • Anything where the goal is to make your vision usable by other people

We don't build your core model. We don't touch your training data. We don't want to see your customer list. We build the part that turns your work into something people can interact with.

How to start

Email Nik at bear@bearbrown.co with a rough description of what you want to build. We'll scope it in a short call, match you with a grad, and start week one.

Nik Bear Brown, PhD · Associate Teaching Professor, College of Engineering, Northeastern University bearbrown.co · nikbearbrown.com