Sandeep Khuperkar

Cofounder & CEO at Data Science Wizards (DSW) - AI & Data Science Solution and consulting company. Open source advocate.

[Data Science Wizards @dswizards]

11 July was outstanding! What an incredible and fulfilling #experience it was for us at the DSW #Enterprise AI Symposium. With the exponential footprint of over 85 #AI professionals in Ireland, we witnessed an evening of #insights and discussions.

[Data Science Wizards @dswizards]

This week in "DSW AI HUB", let's dive into our blog for insights on how UnifyAI simplifies the progression from #AI experimentation to scaled #production, ensuring your enterprise's AI endeavors are as seamless and refined as a culinary masterpiece.

[India in Ireland (Embassy of India, Dublin) @IndiainIreland]

Under "India-Ireland Friendship Lecture Series", Mr Sandeep Khuperkar @sandeeplinux, CEO & Co-Founder, @dswizards will speak on 'Ireland As A Hub For Indian Start-ups' along with an introductory address by Ambassador @AkhileshIFS on Thursday, 16-Feb-2023 @ 12:00 PM (5.30 PM IST)

[Data Science Wizards @dswizards]

In this article, we have discussed the nuts and bolts of the logistic regression model that one can refer to get acquainted with machine learning models.

https://t.co/epkiRx0JOP

#dsw #machinelearning #datascience #learning #algorithms #artificialintelligence #regressionanalysis

[Data Science Wizards @dswizards]

Let’s take a look at how AI is advancing in different domains, including major domains like #bankingandfinance, #healthcare, #agriculture, #education, and so on.

https://t.co/i0QNX30uBW

#dsw #datasciencewizards #datascience #artificialintelligence #ai #deeplearning #data

[Data Science Wizards @dswizards]

Linear regression models are one of the simple and core topics of the data science field which helps everyone to start their journey of model building.

https://t.co/BaWWYqHwrj

#datasciencewizards #ai #ml #datascience #machinelearning #python #artificialintelligence #ai #coding

[Data Science Wizards @dswizards]

Data Science Wizards at Dublin Tech Summit 2022!!

It was a great pleasure and experience to be part of #Europe's one of the largest and fastest-growing B2B technology summits and to interact and exchange views with Tech giants and promising #startups.

[Data Science Wizards @dswizards]

Data Science Wizards wishes you an amazing year ahead with lots of colors, good luck, and success.

Happy Holi 🥰🎉

#holi #happyholi #india #holifestival #festival #colors #colors #festivalofcolors

[Data Science Wizards @dswizards]

Data analysts and data scientists represent two of the most in-demand, high-paying jobs in 2021. The World Economic Forum Future of Jobs Report 2020 listed these roles at number one for increasing demand across industries, followed immediately by AI and machine learning.

[Data Science Wizards @dswizards]

Deep learning is a type of machine learning and artificial intelligence (AI) that imitates the way humans gain certain types of knowledge. Deep learning is an important element of data science, which includes statistics and predictive modeling.

#MachineLearning #DeepLearning

[Amod Malviya @amodm]

A non-technical application of this pattern is in org design, but I'll leave that to your own imagination. Please do share other examples if you found this useful.

End of 🧵

[Amod Malviya @amodm]

Another example (meaningful for engineers) is setting up, tearing down, of cloud resources (e.g. a VM) being a control plane activity, while using the resource is a data plane activity.

This realisation helps you design the technicals & commercials of the system much better.

[Amod Malviya @amodm]

Control-plane, data-plane patterns show up in other places too, e.g. in supply chain technology, all worker units (e.g. picker, delivery executive, etc) are working in the data plane, while "the orchestrator" (in what order should items be picked) is sitting in the control plane.

[Amod Malviya @amodm]

Obviously there are other unaddressed aspects to this example, but hopefully you got an idea of how web2/web3 can co-exist, if we limited blockchains to control plane activities.

[Amod Malviya @amodm]

If we replaced the email itself with something being on-chain (no idea why anyone would want that, but it's a crazy world), the blockchain will have to scale to throughput needs of the global email system, which is highly inefficient, costly and unnecessary.

[Amod Malviya @amodm]

In such a model, my declaring the "rate card" is a control plane activity (I'll probably do it once a month, probably even automatable via the provider), while processing the handling cost per email is a data plane activity.

[Amod Malviya @amodm]

Say, my email provider honours this SMTP extension, and for every incoming email checks against (a cached view of) the latest blockchain DB to see if the declared value by the sender (included as an SMTP header in the email), is higher than the one I'd set or not.

[Amod Malviya @amodm]

e.g. imagine an extension to SMTP (email protocol) where the sender needs to pay to show up in my inbox. I update a list of (sender/domain, cost) tuples, signed by me, into the blockchain, on a regular basis (monthly, quarterly, whatever). I keep high, non-zero costs for spammers

[Amod Malviya @amodm]

Blockchains (at least the truly decentralised ones), require distributed consensus by nature, which, by definition, is going to be noticeably costlier and slower than centralised systems. This is what shows up in the much higher latencies & throughput.

[Amod Malviya @amodm]

Relevance to web3: In many private conversations, I've shared with folks that blockchains are potentially great* for control plane use cases, but not data plane. Let me elaborate.

*(leaving aside a few unaddressed concerns)