Infinite Curiosity Pod with Prateek Joshi

Eran Yahav is the cofounder and CTO of Tabnine, an AI assistant that developers can use to build software faster. He's a professor at Technion - Israel Institute of Technology and was previously a researcher at IBM. He has a PhD in Computer Science from Tel Aviv University.

In this episode, we cover a range of topics including:
- Tasks in software development
- What tasks are likely to benefit from LLMs
- The launch of Tabnine Chat 
- Characteristics of a good AI coding assistant
- Making AI coding assistants context-aware
- Generic LLMs vs domain specific LLMs
- AI copilot for devops work

Eran's favorite book: Catch-22 (Author: Joseph Heller)

--------
Where to find Prateek Joshi:

Newsletter: https://prateekjoshi.substack.com 
Website: https://prateekj.com 
LinkedIn: https://www.linkedin.com/in/prateek-joshi-91047b19 
Twitter: https://twitter.com/prateekvjoshi 

In this episode, the host Prateek Joshi talks about:
- why do we need data orchestration
- what exactly is it
- how does it work
- where is it used in the real world
- market landscape of tools

--------
Where to find Prateek Joshi:

Newsletter: https://prateekjoshi.substack.com 
Website: https://prateekj.com 
LinkedIn: https://www.linkedin.com/in/prateek-joshi-91047b19 
Twitter: https://twitter.com/prateekvjoshi 

Karthik Dinakar is the cofounder and CTO of Pienso, an interactive AI platform anyone can use to turn text data into insights without requiring any coding. He has a PhD in Machine Learning from MIT. He has held roles at MIT Media Lab, Microsoft, and Deutsche Bank.

In this episode, we cover a range of topics including:
- Tech stack of Generative AI applications
- Problem with general purpose LLMs
- Training LLMs on my own data
- Making LLMs faster and cheaper
- Shortage of AI compute
- Context windows
- Pursuit of AGI

Karthik's favorite books: 
- Introduction to Algorithms (Authors: Thomas H. Cormen, Charles E. Leiserson, Ronald L. Rivest, Clifford Stein)
- Mindstorms: Children, Computers, And Powerful Ideas (Author: Seymour Peypert)

--------
Where to find Prateek Joshi:

Newsletter: https://prateekjoshi.substack.com 
Website: https://prateekj.com 
LinkedIn: https://www.linkedin.com/in/prateek-joshi-91047b19 
Twitter: https://twitter.com/prateekvjoshi 

In this episode, the host Prateek Joshi talks about 7 skills that ML builders should develop when they build products.

--------
Where to find Prateek Joshi:

Newsletter: https://prateekjoshi.substack.com 
Website: https://prateekj.com 
LinkedIn: https://www.linkedin.com/in/prateek-joshi-91047b19 
Twitter: https://twitter.com/prateekvjoshi 

Venkat Venkataramani is the cofounder and CEO of Rockset, a real-time search and analytics database built for the cloud. They've raised more than $60M in funding from the likes of Greylock and Sequoia. He was previously at Meta for 8 years and Oracle for 5.5 years. He has spent his entire life building data products, so he has a ton of insights on this topic.

In this episode, we cover a range of topics including:
- Separation of storage and compute
- Isolating compute for ingest and query
- Converged Index
- Vector search
- Stream Processing Systems vs Real-Time Analytics Databases
- Data infra for AI applications
- Bringing compute to data

Venkat's favorite books:
- The Hitchhiker's Guide to the Galaxy (Author: Douglas Adams)
- The Three-Body Problem (Author: Liu Cixin)

--------
Where to find Prateek Joshi:

Newsletter: https://prateekjoshi.substack.com 
Website: https://prateekj.com 
LinkedIn: https://www.linkedin.com/in/prateek-joshi-91047b19 
Twitter: https://twitter.com/prateekvjoshi 

In this episode, the host Prateek Joshi talks about:
- The concept of Geometric Deep Learning
- Why do we need it
- How does it work
- Where is it used in the real world

--------
Where to find Prateek Joshi:

Newsletter: https://prateekjoshi.substack.com 
Website: https://prateekj.com 
LinkedIn: https://www.linkedin.com/in/prateek-joshi-91047b19 
Twitter: https://twitter.com/prateekvjoshi 

Luis Ceze is the cofounder and CEO of OctoML, a platform that offers compute infrastructure to fine-tune, run, and scale your AI models. He's a professor at University of Washington and a venture partner at Madrona. He was previously the cofounder of Corensic. He has a PhD in Computer Science from University of Illinois Urbana-Champaign.

In this episode, we cover a range of topics including:
- OctoAI product announcement
- How to make LLMs faster and cheaper
- Training your own LLMs
- The perceived shortage of AI compute
- Enterprise spend on AI compute
- Applications being built using OctoML
- Domain specific models

Luis's favorite books:
- Thinking, Fast and Slow (Author: Daniel Kahneman)
- Blindness (Author: Jose Saramago)

--------
Where to find Prateek Joshi:

Newsletter: https://prateekjoshi.substack.com 
Website: https://prateekj.com 
LinkedIn: https://www.linkedin.com/in/prateek-joshi-91047b19 
Twitter: https://twitter.com/prateekvjoshi 

Doing customer discovery is a critical skill that founders need to develop in the early stages of their AI startup's journey. In this episode, the host Prateek Joshi talks about 9 steps to do customer discovery for AI products.

--------
Where to find Prateek Joshi:

Newsletter: https://prateekjoshi.substack.com 
Website: https://prateekj.com 
LinkedIn: https://www.linkedin.com/in/prateek-joshi-91047b19 
Twitter: https://twitter.com/prateekvjoshi 

Andre Zayarni is the cofounder and CEO of Qdrant, an open-source vector database company. He was previously the CTO of MoBerries building a next generation HR product based on AI and Real-time. He's been building software products for 20+ years. 

Karthik Dinakar is the cofounder and CTO of Pienso, an interactive AI platform anyone can use to turn text data into insights without requiring any coding. He has a PhD in Machine Learning from MIT. He has held roles at MIT Media Lab, Microsoft, and Deutsche Bank.

In this episode, we cover a range of topics including:
- Why do we need vector databases?
- What type of applications need a vector database?
- When choosing a vector database, what characteristics should a developer look for?
- What are the building blocks of generative AI applications?
- What can LLMs do well and where are the gaps?
- What LLM applications are being built in the enterprise?
- Retrieval Augmented Generation
- What's the importance of vector databases in vertical AI?

--------
Where to find Prateek Joshi:

Newsletter: https://prateekjoshi.substack.com 
Website: https://prateekj.com 
LinkedIn: https://www.linkedin.com/in/prateek-joshi-91047b19 
Twitter: https://twitter.com/prateekvjoshi 

In this episode, the host Prateek Joshi talks about why we need Federated Learning, how it works, and where it's used in the real world.

--------
Where to find Prateek Joshi:

Newsletter: https://prateekjoshi.substack.com 
Website: https://prateekj.com 
LinkedIn: https://www.linkedin.com/in/prateek-joshi-91047b19 
Twitter: https://twitter.com/prateekvjoshi