Excited to present Kamu at @protocollabs.bsky.social Demo Day and make the case that deep-tech innovation in Data actually exists.
Attend online March 16-23: directory.plnetwork.io/demoday/plW26/
Our latest video:
www.youtube.com/watch?v=ZyiZ...
Curious how collaboration on the scale similar to open-source software revolution can be achieved in data?
Check out opendatafabric.org
ODF is a community-driven spec, think "HTTPs of multi-party data exchange" developed by @kamu.dev
Intro video:
www.youtube.com/watch?v=ZyiZ...
Doing BD research in Web3 seriously messes with my belief in overall greater good technology can bring.
Just past two weeks researched 40+ AI / GPU sharing networks, 20+ game streaming DePINs
All have same exact playbook:
1/3
Databricks, Snowflake, Microsoft are stuck in local optima of providing solutions only for internal data efficiencies. Rise of AI exposed acute lack of technology for cross-organizational data exchange. Combination of temporal processing and verifiable computing reveals a new optima that Kamu calls "Global data supply chains".
#WSJ reports that Databricks hit $4B in revenue 🤯
Is it the end-game of enterprise data?
Although a great product, Databricks does NOTHING to facilitate data flows between companies.
In AI era the next frontier of data is *multi-party*:
www.kamu.dev/blog/2024-06...
#dataengineering #bigdata #Web3
Infrastructure of a typical DePIN with high degree of centralization
Using Kamu DePIN negotiates access to personally-owned data and refines it via decentralized pipelines
"Web3 data" is currently synonymous with decentralized file systems. But what about real-time data processing and big data analytics?
It's a very different game in terms of ownership and privacy, but we found a solution.
#web3 #dataengineering #DePIN #blockchaintechnology
In our model if a company behind a DePIN fails - community can preserve all data, or even take over the data pipelines and keep the whole network operational.
This model, is how put "De" back into DePINs and make them behave more like actual DAOs.
#Web3 #DePIN #DataEng
4/4
Solution: Decentralized data middleware where:
- Individual contributors own and control the data they collect
- DePINs access but *don't own* the data
- Data is processed and distributed via transparent and verifiable pipelines
3/4
Pinning raw data in #IPFS or #Arweave is just "Web3-washing" the problem. The true value of a DePIN is in its data collection, verification, quality scoring, and distribution pipelines, which are often closed-source and extremely hard to reconstruct if lost.
2/3
Today, if a DePIN company doesn’t pay its infra bill - the data of entire network is lost, the token value tanks, and years of contributions of people and their hard-earned rewards get wiped out.
Blockchais provide a solution for decentralized money, but decentralization of data is unsolved.
1/3