Gemini-Powered Video Agents Secure $4M Pre-Seed for Fika Jobs

On June 23, 2026, Stockholm-based Fika Jobs secured a $4 million pre-seed round to scale a video-first recruitment platform. Led by Luminar Ventures, the funding backs a system that replaces text resumes with automated candidate profiles generated by conversational AI agents.

Architecture and Agent Interactions

The platform operates as a two-sided marketplace where candidates maintain a persistent video profile rather than submitting distinct manual applications for every role. The ingestion pipeline begins with a LinkedIn profile scrape, which the system uses to dynamically generate personalized interview prompts. Candidates then enter a roughly 10-minute video session with an AI agent powered by Google’s Gemini models.

Following the interview, the platform edits the raw footage into short, discoverable video clips designed to highlight specific professional traits like communication skills, ambition, and grit. For developers building real-time voice agents, this architecture shifts the compute burden compared to standard hiring tools. Existing platforms like Mercor, Maki, and Alex focus on employer-side screening, requiring intensive token processing to parse thousands of applications per role. Fika Jobs moves the AI processing to the candidate side, generating a reusable media artifact that employers can browse directly.

Pilot Adoption and Business Model

During late 2025 and early 2026, the platform entered a pilot phase with more than 50 companies, including Rebtel, Kognity, SICS.ai, and Plenty Labs. The startup currently maintains a waitlist of over 100 employers. Fika Jobs monetizes the resulting placements through a performance-based fee, charging 10% of a successful hire’s first-year salary.

The capital injection includes participation from Alliance VC and Wave Ventures, alongside individual investors including King co-founders Sebastian Knutsson and Riccardo Zacconi. Founded by Jakob and Alexander Dubois, the company will use the funds to expand its engineering team to 10 people ahead of a broader public launch scheduled for the autumn of 2026.

Regulatory and Evaluation Constraints

Deploying conversational AI for hiring introduces strict compliance requirements, particularly under the EU AI Act. Processing raw candidate video requires managing visual and auditory data that can trigger algorithmic bias alerts. If you evaluate and test AI agents for production use, video analysis in a regulated domain like employment necessitates distinct guardrails compared to text-based resume parsers. The underlying model must accurately evaluate soft skills without factoring in protected demographic characteristics visible or audible on camera.

If you are integrating AI into hiring workflows, consider restricting your agent implementations to standardized, reusable candidate profiles rather than continuous employer-side filtering. Moving the AI interaction to the candidate’s initial onboarding reduces redundant API calls and establishes a persistent baseline profile for matching algorithms to query against.