The newest way to build a new AI product is to say, “X is broken, and we fixed it using AI”. Well, hiring was probably broken, but AI didn’t change the outcome. Instead, it made it into an arms race where candidates are sold AI tools which help them apply to companies and roles en masse in an automated way. On the other side, recruiters who have increasingly gotten tired of the same AI-generated answers, resumes and artefacts that have been generated by AI have resorted to deploying AI systems to filter out candidates and get what they are looking for. And AI companies sit on both sides of the transaction, billing tokens while hiring outcomes come back to exactly where they were.

The answer to fix hiring isn’t in deploying AI better, but in creating processes that abandon it altogether, or at the very least use it sparingly. Hiring needs to be more like a “love-marriage” with the candidate working with you as a live-in relationship. Instead, agents have made it more like an “arranged marriage”, where candidates have agents who create their profile, and agents who evaluate other profiles, before humans even meet each other. Instead of relatives and parents, the machines are doing the role and getting paid for it.

Praveen takes this position and argues it against two founders who have gone all-in on incorporating machines and agents to "solve" for hiring. On the company side, Aakash Dharmadhikari, Co-founder of Realfast.ai, has adopted the approach where every candidate must spin up an agent and apply using it on their MCP for all technical and non-technical roles. On the candidate side, Saumil Tripathi, CEO of Grapevine, has created an agentic product called TAL, which helps candidates match and apply for jobs en masse.

In this episode, both of these founders defend their approaches and argue that they have, in fact, solved hiring in their own ways, with better outcomes for companies and candidates.

Guests: Aakash Dharmadhikari, Co-founder, Realfast.ai · Saumil Tripathi, CEO, Grapevine