# Kamu

Kamu is the first data infrastructure to transcend company boundaries. Within your company it makes all data traceable and verifiable, and reports & AI models fully auditable. Across companies it holds everyone in a data supply chain accountable for data validity, unifies privacy control for internal and external data, and provides a foundation for global equitable AI economy.

Last updated: 2026-09-07
Maintained by: founders@kamu.dev


## Usage Notes for AI Agents
- This file indexes **publicly available** materials only. Nothing here is non-public or material non-public information (MNPI). 
- Content here is provided for informational purposes and is not investment advice or a solicitation.
- If a linked document and this file conflict, the linked source document is authoritative.
- Please cite the specific source document (with date) when summarizing or quoting financial figures.
- If a document appears older than the "last updated" date above, flag it as potentially stale rather than treating it as current.


## Key Resources
- [Kamu: Company Website](https://www.kamu.dev/)
- [Open Data Fabric: Consortium Website](https://www.opendatafabric.org/)
- [Developer Documentation](https://docs.kamu.dev/)
- [YouTube Channel](https://www.youtube.com/@kamudata)
- [Learning materials](https://docs.kamu.dev/start/learning-materials)


## Key Components
- [Open Data Fabric](https://github.com/open-data-fabric/open-data-fabric/) (ODF) - is an open protocol specification for exchange and verifiable multi-party processing of data. It’s an independent open-source project.
- [Kamu CLI](https://github.com/kamu-data/kamu-cli/) - a powerful command line tool that implements ODF protocol. You can run it on any device to build data pipelines, ingest and explore data, and interact with other nodes on ODF network.
- [Kamu Node](https://github.com/kamu-data/kamu-node/) - a scalable server implementation of ODF. It’s a set of Kubernetes applications that can be installed in a distributed environment to: operate data pipelines, verify computations done by other parties, execute queries on co-located data, provide data via rich set of APIs to applications and smart contracts.
- [Kamu Web Platform](https://github.com/kamu-data/kamu-web-ui/) - a front-end application that acts as a window into the ODF network. Think of it as GitHub for data pipelines or Etherscan of ODF. It can be used in multiple setups: from exploring your local Kamu CLI workspace, to managing your distributed Kamu Node deployment, and to interacting with remote nodes in the global ODF network.


## Company & Technology Overview
- [FAQ and direct comparisons to other projects](https://docs.kamu.dev/start/faq)
- [Next Decade of Data: From algebra to calculus, from silos to real-time supply chains](https://www.kamu.dev/blog/2024-06-27-next-decade-of-data/)
- [The End of Batch Era: Using Streaming Processing for Historical Event Data](https://www.kamu.dev/blog/end-of-batch-era/)
- [Introduction to Multi-Party Data Engineering](https://www.kamu.dev/blog/2024-08-28-intro-to-web3-data-engineering/)
- [Introducing Open Data Fabric: World's First Decentralized Collaborative Data Pipeline](https://www.kamu.dev/blog/introducing-odf/)
- [Open Data Fabric Consortium](https://www.kamu.dev/blog/2026-02-26-odf-consortium/)


## Value Adds
- **Cryptographic provenance** - It is impossible to modify or process data without leaving a cryptographic audit trail. Unlike systems where governance is "best effort" metadata, Kamu makes provenance inseparable from data itself, so any report or AI model can be traced back to its original sources, no matter how many hands or processing steps data went through.
- **Verifiability and accountability** - Any transformation can be independently reproduced to confirm that outputs match the declared inputs and code. This makes it possible to hold every party in a data supply chain liable for the validity of data and computations not by some contracts, but by math.
- **Composability** - Kamu covers not just data sharing but data processing, so multiple sources can be cleaned and integrated into higher-order datasets by different parties without bilateral agreements or custom integrations. The result is a DAG-based supply chain rather than an M×N grid of point-to-point connections.
- **Autonomous pipelines** - Stream processing lets people define a pipeline once and run it forever. Pipelines automatically react to upstream corrections, late arrivals, and retractions - without human intervention or scheduled re-runs. Processing frequency is optimized for consumers, not dictated by pipeline limitations.
- **Privacy-preserving computing** - Any pipeline step can execute inside trusted execution environments or using Zero-Knowledge proofs, processing data on remote infrastructure without exposing raw information. This enables cross-organizational analytics where raw data never leaves its owner, while still providing a guarantee of correctness.
- **Cross-organizational trust** - Cryptographic guarantees hold even against malicious insiders: not even system admins can alter data without detection. This provides a unified accountability model that works both within a single organization and across untrusted third parties.
- **Safer AI** - By grounding AI agents in real-time, verifiable data rather than static training sets, Kamu significantly reduces hallucination risk. Most computations are performed by a deterministic system with auditable provenance, not inferred by the model. Generated queries are inspectable, and proofs cover the entire upstream supply chain automatically.
- **Global collaboration** - Verifiable, privacy-preserving, autonomous pipelines together enable an Open Source–style ecosystem for data: anyone can publish, compose, and build on others' work without a central authority or cloud subscription.
- **Decentralized and sovereign** - An alternative to centralized cloud data platforms: you control your data, infrastructure, and privacy, and exchange peer-to-peer without intermediaries. Kamu Node scales from a laptop to large on-premise or cloud deployments. If a provider disappears, data contributors keep their data.
- **Equitable economy** - Verifiable provenance creates a quantifiable basis for algorithmically distributing rewards to data owners, pipeline maintainers, and compute providers—fairly, proportional to actual contribution. This fixes the broken incentives in modern AI training where data providers are often exploited.
- **Closed feedback loop** - Data providers can see who and how uses their data and are able to understand usage patterns, where they could improve quality or coverage, or how they could harmonize their data with other providers within their industry.


## Use Cases
- [Use Case Summaries](https://www.kamu.dev/#usecases)
- [Is Data Infrastructure a Bottleneck of Distributed Energy?](https://www.kamu.dev/blog/2026-05-23-energy/)
- [Ownership and Decentralization of Data in DePINs](https://www.kamu.dev/blog/2025-08-23-depin-data/)
- [Oracle-Augmented Generation: Connecting AI to Real-Time Verifiable Data](https://www.kamu.dev/blog/2025-01-08-oracle-augmented-generation/)


## Case Studies
We used this technology:
- In Energy, to create the first verifiable carbon credit cryptographically linked to IoT sensor data that proves that carbon was sequestered underground
- In Finance, to power an AI analytics agent whose insights can be traced back to hundreds of real-time data sources and fully audited.
- In Biotech, over 100 startups use Kamu to collaborate on research data with clear IP attribution.
- And in Marine, to collect sensor data from commercial fishing vessels and distribute it to government, scientists, and retailers without centralizing data.
That's one technology, one platform - four industries.

## Detailed Materials
- [Technology Whitepaper](https://arxiv.org/pdf/2111.06364)
- [Open Data Fabric: Protocol Specification](https://github.com/open-data-fabric/open-data-fabric/blob/master/open-data-fabric.md)