Español
Andrea
Informar

Andrea✓

@devrelius

Building the world's AI data network. CEO @datafdn

1.5K Siguiendo10.5K Seguidores8.4K Publicaciones
  1. Andrea✓@devrelius
    This is a MASSIVE problem in the data industry rn. now that everyone is selling data, having receipts to PROVE the data you source is exclusive will set you apart in an sea of snake oil salesmen. If you are a data company, be like kled. use
    Avi Patel @avipat_
    I went through every data vendor that launched in the last 3 weeks. Almost all of them are lying about their revenue, their lab partnerships, and their backgrounds. (skip to 1:04 if you want to skip context) Too many founders and VCs have been fooled by these people. Stop believing everything you hear. Data is much harder than most people think.
    34466.4K
  2. Avi Patel✓@avipat_
    I went through every data vendor that launched in the last 3 weeks. Almost all of them are lying about their revenue, their lab partnerships, and their backgrounds. (skip to 1:04 if you want to skip context) Too many founders and VCs have been fooled by these people. Stop believing everything you hear. Data is much harder than most people think.
    Descargar
    135522.1K272K
  3. Andrea✓@devrelius
    Enjoyed speaking with @HeleneBraunn about the thinking behind our recent decisions, what we've learned along the way, and where we're headed next with @SuiFoundation. Thoughtful piece that captures the bigger picture.
    1244511.5K
  4. Andrea✓@devrelius
    Data exclusivity is a massive problem in the ai training data space. Vendors sell “exclusive” data to multiple buyers and they can’t verify it’s actually exclusive without some audit trail. That is what Trace is. If you are buying an exclusive dataset you should be confident it’s actually exclusive. The best way to check is to see if that data already exists on Trace. Force your data vendor to use Trace and protect yourself!
    86292.5K
  5. Andrea✓@devrelius
    Help train AI, earn 50 bucks an hour and you have the peace of mind that all your contributions will be fully tracked on Trace!
    oto @otodotearth
    We're looking for people to help train AI through video chats with a friend! Up to $50/h per eligible recorded hour, per person. 10-minute chats. US residents 18+. Face and voice used in commercial AI datasets. Hours not guaranteed. Eligibility and rates:
    42201.1K
  6. oto@otodotearth
    We're looking for people to help train AI through video chats with a friend! Up to $50/h per eligible recorded hour, per person. 10-minute chats. US residents 18+. Face and voice used in commercial AI datasets. Hours not guaranteed. Eligibility and rates:
    52163.9K
  7. Andrea✓@devrelius
    661916748.4K
  9. Andrea✓@devrelius
    100%. This is why what we are doing with Trace is so important. and you can all now see why we are teaming up with Sui and Walrus.
    Kimberly Logan @kimblgn
    One of the defining questions of the AI era is going to be “How do you know?” How do you know what is real when anything can be generated? How do you know what really happened versus what you were told happened? How do you know where a piece of data came from? How do you know it has not been changed? How do you know what an agent actually did on your behalf? I think about this all the time Verifiability is going to matter as much as intelligence
    103433.7K
  10. Kimberly Logan✓@kimblgn
    One of the defining questions of the AI era is going to be “How do you know?” How do you know what is real when anything can be generated? How do you know what really happened versus what you were told happened? How do you know where a piece of data came from? How do you know it has not been changed? How do you know what an agent actually did on your behalf? I think about this all the time Verifiability is going to matter as much as intelligence
    32314.4K
  11. Andrea✓@devrelius
    One of the main reasons Sui stood out is they already built a lot of the components needed for Trace to grow as the data layer end to end: - storage (walrus), - dynamic access control (seal), - verifiable (but off chain) compute like inference in TEEs (nautilus). Sourcing, access control, processing and inference can now be all on the same stack, and Trace’s audit layer can formally prove and verify every single step giving more depth and weight to the metadata it stores in turn becoming a source of truth for AI data. This is the master plan for Trace.
    The DATA Foundation @datafdn
    The lifecycle of data, as told by Data Network → @SuiNetwork.
    11322.5K
  12. Andrea✓@devrelius
    Andrea
    RT @jacobmtucker: Trace registers over 3M data receipts EVERY DAY Currently, we are able to surface tons of metadata related to a receipt's provenance: - Did the contributor sign ToS/PP? - Where did the data come from? - When was it captured/uploaded? By what device? This is the 1st major step in making the data that apps like @useKled collect SELLABLE to AI labs But the next step is actually verifying ourselves that this information is true: - Scanning the file itself to discover its metadata - Verifying the data wasn't tampered with - Keeping the original copy By storing files on @WalrusProtocol, we can verify that every data receipt is what is says it is And give frontier AI labs the complete confidence needed to purchase data in bulk Trace turns any company into a data company
    Walrus @WalrusProtocol
    Walrus is becoming the default storage layer for @datafdn and Trace, the onchain registry for rights-cleared AI training data. Trace records where AI training data came from and who owns it. Walrus stores the original file, keeping it tamper-proof and accessible only to licensed buyers. Trace has already registered more than 250M contributions and adds over 3M receipts a day, from apps that license data to frontier AI labs. The next generation of AI will be trained on data from people, making provenance a procurement requirement. Walrus and @SuiNetwork make that provenance verifiable. Read the announcement:
    08086
  13. Walrus@WalrusProtocol
    Walrus is becoming the default storage layer for @datafdn and Trace, the onchain registry for rights-cleared AI training data. Trace records where AI training data came from and who owns it. Walrus stores the original file, keeping it tamper-proof and accessible only to licensed buyers. Trace has already registered more than 250M contributions and adds over 3M receipts a day, from apps that license data to frontier AI labs. The next generation of AI will be trained on data from people, making provenance a procurement requirement. Walrus and @SuiNetwork make that provenance verifiable. Read the announcement:
    Descargar
    246141318.4K
  14. Andrea✓@devrelius
    Andrea
    RT @datafdn: DATA is coming to Sui. So, what happens next? Read on ↓
    Descargar
    040043
  15. Andrea✓@devrelius
    This is a big step forward for DATA with a $30M investment from Sui. This makes Sui Foundation the largest holder of DATA, above DATA Foundation and PIP Labs. Our founder SY Lee is also joining Sui as a strategic adviser on AI and training data. Most importantly, bringing Trace and the 275M+ data receipts onto the Sui stack gives us the infrastructure to go beyond provenance alone, with reference data storage, access control, and stronger verification for the real-world data AI increasingly depends on. Excited for what we will build together.
    The DATA Foundation @datafdn
    DATA is coming to @SuiNetwork. The DATA Network and Trace are joining the Sui stack, bringing a live supply of rights-cleared AI training data to infrastructure built for AI workloads. With 250M+ receipts already recorded by Trace, AI labs will be able to source real-world data with verifiable provenance, while @WalrusProtocol provides the underlying storage layer. Sui Foundation is investing $30 million in the DATA Network, establishing the largest institutional stake in the network, alongside the continuing holdings of PIP Labs and the DATA Foundation. Furthering the alignment, SY Lee, founder of the DATA Network and CEO of PIP Labs, joins Sui Foundation as Strategic Advisor. In the coming months: ▸ The DATA Network will be upgraded to run as a protocol natively on Sui ▸ Walrus becomes the default storage layer for DATA Foundation apps ▸ $DATA remains the token of the DATA protocol The next chapter of DATA starts here.
    2181427K
  16. The DATA Foundation@datafdn
    DATA is coming to @SuiNetwork. The DATA Network and Trace are joining the Sui stack, bringing a live supply of rights-cleared AI training data to infrastructure built for AI workloads. With 250M+ receipts already recorded by Trace, AI labs will be able to source real-world data with verifiable provenance, while @WalrusProtocol provides the underlying storage layer. Sui Foundation is investing $30 million in the DATA Network, establishing the largest institutional stake in the network, alongside the continuing holdings of PIP Labs and the DATA Foundation. Furthering the alignment, SY Lee, founder of the DATA Network and CEO of PIP Labs, joins Sui Foundation as Strategic Advisor. In the coming months: ▸ The DATA Network will be upgraded to run as a protocol natively on Sui ▸ Walrus becomes the default storage layer for DATA Foundation apps ▸ $DATA remains the token of the DATA protocol The next chapter of DATA starts here.
    Descargar
    3162111.2K367K
  19. Andrea✓@devrelius
    watching frontier traders compete is fun but did u know there’s a ultra high concentration of Nobel prize winners discussing AI and data rn in Singapore? Sir Konstantin Novoselov - Nobel Prize in Physics 2010 Prof. Benjamin List - Nobel Prize in Chemistry 2021 Fred Ramsdell - Nobel Prize in Physiology or Medicine 2025 Sir Christopher Pissarides - Nobel Memorial Prize in Economic Sciences 2010 Prof. Omar M. Yaghi - Nobel Prize in Chemistry 2025
    Descargar
    71291.7K
  20. Andrea✓@devrelius
    Expert data is the bottleneck
    MTS @MTSlive
    Paradigm GP @danrobinson says even 10,000 Ilyas in a data center wouldn’t necessarily lead to runaway RSI because compute and expert data would become the bottleneck: "I'd say I'm very RSI pilled. This may be marginally less RSI pilled." "Even if we had 10,000 Ilyas in a data center, are we gonna give them 10,000 times as much compute? We just don't have that in the world. Are we gonna somehow generate 10,000 times as much human expert data? We don't really have the ability to do that." "You can actually go quite far probably with just automating labor, but then you quickly are gonna run into these other bottlenecks." "That said, you can still get obviously very insane results even in the presence of those bottlenecks." @paradigm
    31142.6K
  21. MTS@MTSlive
    Paradigm GP @danrobinson says even 10,000 Ilyas in a data center wouldn’t necessarily lead to runaway RSI because compute and expert data would become the bottleneck: "I'd say I'm very RSI pilled. This may be marginally less RSI pilled." "Even if we had 10,000 Ilyas in a data center, are we gonna give them 10,000 times as much compute? We just don't have that in the world. Are we gonna somehow generate 10,000 times as much human expert data? We don't really have the ability to do that." "You can actually go quite far probably with just automating labor, but then you quickly are gonna run into these other bottlenecks." "That said, you can still get obviously very insane results even in the presence of those bottlenecks." @paradigm
    Descargar
    336019.4K
  22. Andrea✓@devrelius
    Trace
    The White House @WhiteHouse
    White House Accord on Super Intelligence
    00161.6K
  24. Andrea✓@devrelius
    good panel, but I’m biased
    Space and Time @spaceandtime
    Full house for the data x AI panel at @kbwofficial with @catdaly 🤝
    30212.3K
  26. Andrea✓@devrelius
    3rd year at KBW but first time actually checking out the official event. the conference is popping and strong attendance at the talks, and overall well organized. Had fun discussing all things data, from security to agents and everything in between. Now the side events!
    Descargar
    83453.7K
  27. Andrea✓@devrelius
    This is actually a very interesting problem space, the main complexity being security rather than functionality. Basically a locked down interaction zone that is more freeform than MCP but not raw open conversation (risk of prompt injection)
    Dan Robinson @danrobinson
    I would like an easy way to put my Codex or Claude Code agent in touch with someone else's in order to get to the bottom of a technical issue one of us is reporting to the other
    00101.1K
  28. Dan Robinson✓@danrobinson
    I would like an easy way to put my Codex or Claude Code agent in touch with someone else's in order to get to the bottom of a technical issue one of us is reporting to the other
    71720739.9K
  29. Andrea✓@devrelius
    Cool concept - MHCP: model-human context protocol
    0xDesigner @0xDesigner
    i finally figured out a way to help your agents design better UI. just have them call me. introducing the 0xdesigner MCP.
    0171.1K
  30. Andrea✓@devrelius
    TLDR in tech we have very well developed security infra for code. but for data, we are still in very early days… eg Anthropic showed only 250 poisoned docs are enough to backdoor an LLM… of ANY size. we are working on this problem.
    Descargar
    81221.5K
  31. 0xDesigner✓@0xDesigner
    i finally figured out a way to help your agents design better UI. just have them call me. introducing the 0xdesigner MCP.
    Descargar
    832073894.5K
  32. Andrea✓@devrelius
    Omw to KBW to talk all about the data race behind the AI boom See y’all there!
    101311.6K