@pateljm Profile picture

Jignesh Patel

@pateljm

Professor, Co-founder and CEO DataChat

Similar User
Wisconsin DB Group photo

@wiscdb

ACM SIGMOD photo

@sigmod

CMU Database Group photo

@CMUDB

Sam Madden photo

@samrmadden

Andreas Kipf photo

@andreaskipf

Ippokratis Pandis photo

@Hippotas

Alvin Cheung photo

@alvinkcheung

Dutch Seminar on Data Systems Design photo

@dsdsdnl

Badrish Chandramouli photo

@badrishc

Immanuel Trummer photo

@ImmanuelTrummer

Suyash Gupta photo

@suyash_sg

Natacha Crooks photo

@siobhcroo

Pınar Tözün (@pinar@social.itu.dk) photo

@pinartozun

Theo Rekatsinas photo

@thodrek

Joe Hellerstein photo

@joe_hellerstein

Shasank Chavan from @Oracle gave a fantastic talk at @CMUDB, describing Oracle's new vector search feature and how it integrates with their transaction and scale-out capabilities. Seeing the SQL extensions and vector indexing in action was also exciting. youtu.be/byo2qJbnQz0


Nikhil Shah, Bowei Chen, @sortalongo, Olivia Zhou, and Eli Ruiz visited CMU in September. Here’s Dan describing the planned "finalize" syntax for continuous query processing (CQP). Check out the new approach to CQP planned for @SnowflakeDB youtu.be/7wiuIRBPNmg


youtu.be/CmgCECKgOo0 Last week, @CMUDB hosted @dbinmemory, Beda H., and Ajit M, who described a new feature in @Oracle 23ai that mixes the JSON and Relational data models. That is 1 EVP and 2 VPs presenting this work -- worth checking out!


A8: It is hard to think of any sector that won't be impacted. But, GenAI will significantly advance the field of medicine (look at the fantastic progress in protein folding), manufacturing (robotics is enabled in dramatic ways with GenAI), and education. #eweekchat


A7: It is super essential for organizations to focus on retraining their workforce to work with GenAI. Retraining takes time, but the competitive advantage comes from having the right people AND GenAI. #eweekchat


A6: Definitely a huge concern, and ignoring can be a disaster. There is a broad range of security concerns, including malicious theft of your data, especially with agent architectures. All this is orthogonal to privacy, which is the other dimension. #eweekchat


A5: Everyone wants to do something with GenAI. In some cases, the use case may not be as compelling to justify the cost. The costs are coming down, though, and soon it will be hard to find any business that is not using it directly or indirectly. #eweekchat


A4: Put in an A/B testing framework upfront so that you know when there is a regression. These models are "finicky," and small changes in how they are invoked (say in your RAG pipeline) have a big impact. You need to know of regressions as soon as they happen. #eWEEKchat


A2: Use it all the time, and on some days, more than a regular web search, which is crazy. It dramatically reduces the time it takes to get started on some idea, but the hard steps, which are often inherently human and related to judgment is still critical #eWEEKchat


It will completely reshape business and, more importantly, how humans work in every job skill. There is a lot of hype, but it is justifiable IMO. Most projections will come through, though not all. But expect even some far-out ideas to become real. #eweekchat


Devin Petersohn from @SnowflakeDB talked to @CMUDB yesterday about how dataframes are meeting databases. If you missed it, you can catch up here: youtu.be/7TyIjqvfWto?fe… Lots of exciting technical nuggets in the talk. Another interesting part is Devin talking about hanging out…


The @datachatAI booth (#330) is now fully set up @Googlecloud Next. We will also present at the NVIDIA booth. Thanks #GoogleCloudNext #NVIDIAInception! Also, we are now in GCP's Advantage Program and on the GCP Marketplace: businesswire.com/news/home/2024…

Tweet Image 1

@datachatAI⁩ booth ⁦@Gartner_inc⁩ is now open. Stop by if you are around.

Tweet Image 1

Loading...

Something went wrong.


Something went wrong.