Recruiterflow vs Folk

Every entry in the window is marketing content, published in same-day batches of two and three. The recurring argument is a definition fight over what AI-native recruiting software means - that bolted-on summarize buttons and copilots do not qualify - now extended into an evaluation framework telling buyers how to see past a demo. The newest post breaks the pattern slightly by naming a capability: AIRA Search, pitched as natural language candidate search that reads recruiting intent and explains why each candidate fits, replacing Boolean strings and filters.

The content is doing competitive work rather than announcing product work - setting evaluation criteria a buyer will carry into a demo, and citing placement-per-recruiter and revenue-decline figures to argue that headcount-light firms are the ones scaling. AIRA, the agent layer inside the product, has moved from inventory and buyer-guide framing to a named search capability with a stated behaviour, which is the criteria set by the evaluation post starting to reappear as product claims. What the feed still will not show is when any of it shipped, since the bodies are teasers that end mid-sentence.

Expect the remaining evaluation criteria from that framework - explaining its reasoning, acting across the whole record rather than one field - to be attached to further named AIRA capabilities. Whether those are shipping or being previewed is not something this source can settle.