SaaS giants vs frontier labs: who's hiring more AI engineers?
The default 2024 narrative — “frontier labs build the models, app companies wire them in” — was true for a year. The hiring data says it’s no longer the shape of the market. Adobe alone is opening more AI engineering roles than every named frontier lab.
What we’re counting
Active job postings classified as AI-related (LLM/ML/agent/data, score ≥ 7), pulled live from each company’s official hiring page once a week. No press-release numbers, no self-reported headcount — only roles that are publicly accepting applications right now. The “frontier labs” cohort is the eight US-accessible labs whose business is training and serving foundation models. The “SaaS pivot” cohort is the best-known 2010s-era SaaS companies that have rebuilt around AI as the primary product surface (not a side feature). Both lists are hand-curated; see methodology for inclusion rules.
The implication
Distribution beats invention, again. Adobe, Salesforce, and ServiceNow already sit inside every Fortune 500 procurement contract. Wiring AI into their core surfaces is the highest-leverage move available to them, and the hiring data shows they’re betting it — not waiting for a startup to disrupt them. The frontier labs still set the upper bound on capability, but the AI engineers shipping value into customers’ workflows increasingly sit at the application companies, not the labs.
Want to go deeper?
Click any company name above to see its full active hiring page — stage breakdown, comp bands, recent narrative shifts. Or compare across the whole tracked universe at /companies. Data refreshes weekly when the cron rebuilds the marts. More on methodology →
