The first wave of AI recruiting tools screened resumes; the second wave is conducting the actual interviews and you're not invited to watch.

The Summary

  • Fika Jobs raised $4M to build a platform where AI agents conduct video interviews with job candidates, then package those interactions into TikTok-style short-form profiles
  • The Stockholm startup is betting that video-first talent discovery beats resume parsing, especially for roles where personality and communication matter more than credential signaling
  • Real signal: This isn't AI helping recruiters work faster. It's AI replacing the entire first-contact layer of hiring, from screening to interview to candidate presentation.

The Signal

Fika Jobs is doing something most hiring AI hasn't touched yet: conducting the actual conversation. Most HR tech uses agents for resume screening, scheduling, or candidate outreach. The platform lets AI agents interview candidates on video, ask follow-up questions, and then create highlight reels that hiring managers can scroll through like social media feeds. The candidate talks to a bot. The hiring manager watches clips. No human interaction required until someone decides to move forward.

The TikTok comparison matters more than it sounds. Short-form video killed the album. It's killing the long-form blog post. Now it's coming for the one-hour screening call. Hiring managers don't have time to watch 30-minute interviews for 50 candidates. But they'll watch 90-second highlight clips that show how someone thinks under pressure, how they explain technical concepts, whether they're charismatic or dead-eyed.

"The first wave of AI recruiting tools screened resumes; the second wave is conducting the actual interviews."

The economics are obvious. A senior recruiter conducting phone screens might handle 10-15 candidates per day if they're moving fast. An AI agent handles hundreds simultaneously, never gets tired, asks consistent questions, and doesn't have unconscious bias about accents or college names. The $4M round suggests investors see a wedge into the $200B+ global recruiting market, starting with roles where personality and communication skills are harder to assess from a resume.

But here's what the company isn't saying: This model works better for some jobs than others. Software engineer? Maybe the AI can assess problem-solving in real-time. VP of Sales? The bot conversation itself becomes the sales test. Entry-level customer service? Perfect fit. But try using an AI agent to interview a senior executive or evaluate cultural fit for a tight-knit startup team, and you hit the ceiling fast.

Key questions this raises:

  • What happens to the 300,000+ corporate recruiters in the US whose job is primarily conducting these first-round screens?
  • Do candidates perform better or worse when they know they're talking to an agent instead of a human?
  • Who's liable when the AI asks an illegal interview question or makes a discriminatory assessment?

The deeper shift: hiring is becoming asynchronous. Candidates record video responses on their schedule. AI processes them on its schedule. Hiring managers review clips on their schedule. Nobody's on the phone at the same time anymore. That's either more humane (no taking calls during your current job) or more dystopian (you're performing for an algorithm, not a person), depending on which side of the labor market you're on.

The Implication

If you're job hunting in 2026, expect to talk to more bots and fewer humans in your first interview. Practice like you're making a YouTube video, not having a conversation. The AI isn't evaluating your resume anymore. It's evaluating your performance.

If you're hiring, the question isn't whether to use AI agents for screening. It's when. Tools like Fika Jobs will be table stakes within 18 months for any company hiring at volume. The competitive advantage shifts to what happens after the AI screen: how quickly you can move candidates through the human stages, how well you close offers, whether your culture actually matches what the highlight reel suggested.

Sources

TechCrunch AI