InfiniMind: Turning Dark Video Data into Business Insights

Unused video is a valuable but overlooked resource

Businesses generate more video than ever. From decades of broadcast archives to thousands of store cameras and hours of production footage, most of it sits unused on servers—unwatched and unanalyzed. This so-called dark data represents a massive, untapped resource: video and audio companies collect automatically but rarely use in meaningful ways.

Founders and company mission

To tackle this problem, Aza Kai (CEO) and Hiraku Yanagita (COO), two former Googlers who worked together for nearly a decade at Google Japan, co-founded InfiniMind. The Tokyo-based startup builds infrastructure that converts petabytes of unviewed video and audio into structured, queryable business data.

"By 2024 the technology matured and market demand was clear, so we felt compelled to build the company ourselves." — Aza Kai

What makes InfiniMind different

Early video tools could label objects in single frames. They struggled to track narratives, understand causality, or answer complex questions across long-form footage. For organizations with decades of archives, even basic queries about content often went unanswered.

InfiniMind leverages recent advances in vision-language models and multimodal AI. Progress between 2021 and 2023 moved video intelligence beyond simple object tagging to understanding sequences, speaker identity, audio context, and causal relationships across scenes. Falling GPU costs and steady performance gains also made large-scale processing feasible.

Key features

• No-code deployment: clients bring their data and get actionable insights without writing code.
• Multimodal analysis: integrates video, audio, speech and sound understanding.
• Scalability: designed to handle long-form content and petabytes of footage efficiently.
• Cost efficiency: optimized to reduce compute and storage costs for enterprises.

Products and traction

InfiniMind launched its first product, TV Pulse, in Japan in April 2025. The AI-powered platform analyzes television content in real time to help media and retail companies track product exposure, brand presence, customer sentiment, and PR impact. After pilot programs with major broadcasters and agencies, TV Pulse already has paying customers, including wholesalers and media firms.

The company's flagship product, DeepFrame, is a long-form video intelligence platform. DeepFrame can process 200 hours of footage to pinpoint specific scenes, speakers, or events. A beta release is scheduled for March, with a full launch in April 2026.

Funding, expansion, and strategy

InfiniMind recently raised $5.8 million in seed funding led by UTEC and joined by CX2, Headline Asia, Chiba Dojo, and an AI researcher associated with a16z Scout. The team plans to relocate its headquarters to the U.S. while maintaining a Japan office. Japan served as an ideal testbed thanks to strong hardware, engineering talent, and a supportive startup ecosystem.

Seed funding will accelerate DeepFrame development, expand engineering infrastructure, hire more engineers, and grow the customer base across Japan and the U.S.

Market positioning and competitors

The video analysis market is fragmented. General-purpose providers such as TwelveLabs offer broad video understanding APIs for consumers, prosumers, and enterprises. InfiniMind focuses on enterprise use cases: monitoring, safety, security, compliance, and deep content insights for broadcasters and retailers.

InfiniMind emphasizes practical tradeoffs: high accuracy paired with cost efficiency and the ability to analyze unlimited video length. This combination aims to solve pain points that many existing tools leave unaddressed.

Use cases and examples

• Retail: analyze store camera footage to measure product placement effectiveness and customer flow.
• Broadcast media: index decades of archives to surface specific scenes, talent appearances, or ad exposure across programs.
• Security and safety: monitor live feeds for anomalies and trigger alerts on safety events.
• PR and marketing: quantify brand visibility and sentiment across TV, streaming, and recorded content.

Vision and future outlook

"Understanding general video intelligence is about understanding reality," Kai said. He framed industrial applications as immediate priorities but described the broader mission as pushing the boundaries of video and multimodal AI to help humans make better decisions.

Why this matters to a broad audience

As video becomes central to business operations and storytelling, tools that unlock dark data will reshape how companies measure performance, manage risk, and build products. For readers interested in related fields—such as gaming news, media analytics, or AI-driven content moderation—this technology signals a shift toward richer, searchable multimedia intelligence.

Image credits: InfiniMind

Event note

TechCrunch event — Boston, MA | June 23, 2026