World's first planet-scale data pipeline
Kamu brings the power of enterprise data pipelines into global trustless environment:
It allows people to collaborate on real-time data processing
Ensures that all data stays verifiably trustworthy
Brings properties of Smart Contracts to SQL, achieving verifiability and accountability levels of blockchains, but at a fraction of the cost
Works with any storage, decentralized (e.g. IPFS) and conventional (e.g. S3)
We exist to empower micro data publishers, fix issues plaguing the data sciences, and bring collaboration to data on the scale as seen today in Software.Watch Demo Try It Yourself Follow our news
We shape a better future of data management
Data is our history - it deserves a better treatment!
Complete Historical Account
Most data solutions nowadays store only the current state of data. They allow updating and deleting data, effectively erasing the history of changes.
In Kamu we believe that data is our history, so we designed the entire system to never forget:
- All data is immutabile and serves as a complete historical record of events
- You can rewind time on any dataset with per-event granularity
- You can use temporal data joins to correlate data over time
Aggregates are great for making sense of large volumes of data, but all aggregated data becomes stale right after it's put together.
In Kamu aggregates are produced not by copying data, but via pure functional transformations:
- We encourage data to be published in a raw unaggregated form
- All updates to raw data are instantly reflected in all aggregates
- An aggregate created once continues to receive updates from its source
Lineage & Provenance
Modern data science promotes a harmful cycle of downloading, modifying, and re-publishing the data. Data becomes disassociated from its origin, it is impossible to know where it came from, and which transformations were applied - such data cannot be trusted.
Kamu keeps a complete record of all data transformations:
- Tell exactly which datasets were used as a base to produce certain aggregated data
- See which transformations were applied
- Trace any piece of data, down to a single cell, back to its origins
The current reality is that organizations that produce valuable data often don't have the right incentives, resources, or skills to provide high quality data. This results in data exposed with sub-optimal frequency, in a variety of poor formats, on a poor infrastructure.
Kamu on the contrary:
- Uses the de-facto standard analytical data formats
- Supports variety of data types, including geospatial data
- Stops malformed data from propagating at the earliest possible stage
- If bad data finds its way in - collaborate with others to make it better