Research

Domain knowledge is what ships AI.

Frontier models are a commodity. What separates a working AI deployment from a stalled pilot is whether the engineers around the model understand the customer's industry deeply enough to bend the product around it. That is why StarPlan is a talent network of industry-specialist AI engineers — not a generic ML marketplace.

6
Research essays
35
Industries we tag
150
Companies searching the network
510
Engineers in the talent pool
Our position

Why we hire AI engineers by industry, not by model.

The model layer can be swapped out in an afternoon. The industry layer — what a regulator will actually accept in healthcare, how a logistics dispatcher really triages exceptions, which fields in an underwriting packet are load-bearing — takes years to acquire and is what customers actually pay for. An engineer with deep exposure to one industry and a weaker model will outship a generalist with the strongest model on every metric the customer cares about: time-to-trust, error rate on the long tail, and operational fit.

That is the entire reason StarPlan is structured the way it is. Every profile carries an industry tag drawn from where the engineer has actually shipped — not where they have read a blog post. Companies filter by their own industry first and the right five to ten people surface immediately, instead of the right five to ten hundred.

More research

Essays on why domain knowledge is the moat in vertical AI, what an enterprise ontology really is, and why the engineers who can build one inside a specific industry are the most valuable seat on an AI team.

Hire AI engineers who already know your industry.

Filter the marketplace by the industry you actually operate in — healthcare, fintech, logistics, legal, and 31 more — and reach the people who have shipped in it directly.