P

Poseidon

Real-world data for physical AI

Performances

RD Popularity Index

408.11% (24h)
0100200300400500
Project Rank: 735/15432

RD Growth Index

3480.43% (24h)
0400800120016002000
Project Rank: 1202/15432

Comparison

Fundraising

Follow Updates

Follow Lists

People

About Poseidon

Poseidon is a decentralized data layer focused on collecting real-world, long-tail data for AI training. Its core business is sourcing, refining, and licensing high-quality datasets—such as voice, image, and video—that are difficult to obtain from public internet sources. The company operates Numo, a mobile app that allows individuals to contribute first-person data (recorded in real environments) and earn rewards directly tied to their contributions.

The market pain point Poseidon addresses is the growing scarcity of diverse, physical-world data needed for embodied AI (robotics, autonomous vehicles, etc.), combined with legal ambiguity around data provenance and usage rights. Poseidon uses blockchain infrastructure to provide verifiable traceability for every data record, ensuring contributors can see how their work is used and compensated, while buyers gain clarity on data origin and licensing.

In the past six months (March–September 2026), Poseidon expanded into South Korea through a partnership with Toss, a major mobile finance platform, bringing Numo to Toss's user base of approximately 30 million people. The collaboration focuses on building Korean-language training data, with Toss users contributing voice, image, and video content. This marks a significant step in validating Poseidon's model in a new market before potential global expansion.

No major negative events were reported in the period. However, the broader industry faces ongoing challenges around data privacy and fair compensation for contributors, which Poseidon aims to address through transparent tracking and payment mechanisms.

Updated: Sep 1, 2026

Poseidon has completed a $15 million seed round led by a16z crypto, marking its formal entry into the decentralized AI data market.

Following this, the project released a lightweight white paper outlining its full-stack architecture, which integrates decentralized data collection, curation, and IP licensing built on Story's infrastructure.

Updated: Sep 1, 2026

Over the past six months, Poseidon has emerged as a full-stack decentralized data layer bridging the gap between data supply and AI demand. Key milestones include a $15M seed round led by a16z, and a strategic partnership with Toss, bringing 30 million users into the AI data economy. The contributor app Numo has recorded over 711,000 data registrations globally, now integrated within the Toss app. Poseidon focuses on sourcing, refining, and licensing IP-safe, composable training datasets, with an emphasis on physical AI data (robotics, autonomous vehicles). Future roadmap includes expanding from Korean market to global markets, extending data types to audio, biometric, and healthcare, and building a provenance layer via The DATA Network to ensure transparent data lineage and compliant licensing.

Updated: Sep 1, 2026

Sandeep Chinchali (Chief Scientist): PhD in Computer Science from Stanford, professor at UT Austin, leading research in edge computing, networked robotics, and generative AI. Previously contributed to Story Protocol as Chief AI Officer.

Sarick (Co-founder): Product-driven AI engineer, built AI systems at LotusFlare and Roadz, former Lead AI Engineer at Story Protocol, specializing in causal inference and multi-agent systems.

LJW (VP of Strategy & Operations): Previously served as Global Strategy Officer at Kakao Entertainment, CEO of Radish Fiction (acquired for $440M), and President of the Oxford Union. Recognized as Forbes 30 Under 30 Asia and Trilateral Commission David Rockefeller Fellow.

Updated: Sep 1, 2026