Hiring threads show AI infrastructure roles dominating demand
The August 2026 'Who is hiring' thread is live, and scanning the roles reveals a clear pattern. PostHog is hiring an AI Research Engineer. Shepherd is hiring for autonomous underwriting. SmarterDx is hiring across multiple roles in AI-assisted clinical documentation. Mbodi AI is hiring robotics and ML engineers. Manufact is hiring for an MCP cloud. The through-line is AI infrastructure and applied ML, not generic software engineering.
The YC hiring board is similarly weighted. Flick is building 'Figma for AI filmmaking.' Trellis AI is building agents for healthcare access. AccessOwl is hiring an AI TypeScript engineer to connect 300 SaaS tools. Even companies not primarily AI-focused are advertising AI-adjacent roles.
The counterpoint is that these roles are genuinely hard to fill. The skill set they want, applied ML with systems engineering depth, is rare. Companies advertising for 'AI Research Engineer' at a SaaS company are often competing against labs for the same small pool of people.
So what?
If you are hiring for AI roles right now, you are competing against well-funded labs and established platforms. Be specific about what 'AI engineer' means at your company, show the technical depth of the problem, and make the compensation visible upfront. Vague listings for 'AI engineers' are getting ignored by the candidates you actually want.