Shofo, a San Francisco-based AI infrastructure startup building large-scale video datasets for machine learning, has raised funding through a combination of accelerator backing and early-stage investment support as it develops what it describes as a “live index” of short-form social media content for AI training. Founded in 2025, the company focuses on collecting, cleaning, segmenting, and labeling millions of hours of video data across platforms such as TikTok, Instagram, X, and LinkedIn, enabling AI labs to access structured datasets for model training and semantic search applications.

The company’s latest disclosed financing activity includes participation from the accelerator program operated by Y Combinator, which has supported Shofo as part of its Winter 2026 cohort. According to multiple startup intelligence profiles, Shofo’s most recent funding is structured as an early-stage convertible note combined with accelerator investment support, reflecting a typical pre-seed to seed-stage setup for companies emerging from YC’s program. The startup reportedly raised approximately $500,000 in its latest disclosed round, with additional non-dilutive program backing also attributed to YC participation.

Shofo operates in the rapidly expanding AI data infrastructure sector, where demand for high-quality training datasets has increased alongside the growth of multimodal foundation models. The company’s core product is a continuously updated index of video content, which it processes into structured datasets that can be queried by AI researchers and developers. This approach positions Shofo as an infrastructure layer between raw social media content and machine learning pipelines, a segment increasingly attracting venture interest due to the bottleneck of labeled training data.

Investor participation is closely tied to Shofo’s accelerator pathway, with Y Combinator serving as both a funding source and a strategic support platform. While detailed angel or institutional investor breakdowns have not been publicly disclosed beyond accelerator involvement, YC’s standard model typically provides capital alongside mentorship and network access, helping startups refine product-market fit and prepare for subsequent venture rounds.

The company was founded by Bryan Hong, Andre Braga, Braiden Dishman, and Alexzendor Misra, who collectively built Shofo to solve the problem of fragmented and inaccessible social media data for AI training. The founders describe the platform as a system that can extract targeted datasets—such as cooking videos featuring hand-object interactions—and deliver them in structured formats suitable for research and model training workflows.

Shofo’s emergence comes at a time when AI labs are increasingly constrained by the availability of high-quality, legally usable training data. By focusing on video-based datasets rather than static text corpora, the company is targeting one of the most computationally and operationally complex areas of AI infrastructure. Its semantic indexing system is designed to allow researchers to query billions of video segments and retrieve precisely filtered subsets of data, reducing the time and cost associated with manual dataset construction.

Although Shofo’s current funding remains relatively modest in scale compared to later-stage AI infrastructure companies, its backing by Y Combinator places it within a pipeline that has historically produced high-growth companies in data tooling and AI infrastructure. The accelerator’s involvement often signals potential for follow-on financing from venture capital firms after demo day, depending on traction, enterprise adoption, and dataset demand from AI labs.

As Shofo continues to expand its video indexing and dataset generation capabilities, its funding trajectory is expected to evolve beyond accelerator support into institutional venture capital rounds. For now, its disclosed capital base and YC affiliation form the core of its financing profile, supporting early product development and infrastructure scaling as it builds toward broader commercialization in the AI data market.

Related Articles