How AI Happens

• The rapid evolution of technology excites Anthony, especially in AI.
• User preferences are shifting towards more human-like AI interactions.
• Empathy in AI is crucial for better customer service experiences.
• The partnership between Amdocs and NVIDIA emphasizes the importance of software efficiency
• Software and hardware advancements must progress in parallel to maximize productivity.
• Physical AI integration will enhance daily life through automation and smart devices.
• Emergent behavior in AI represents a new frontier in reasoning and decision-making.
• Generative AI can learn and adapt beyond traditional if-then programming.
• An audit trail is essential for transparency in AI decision-making processes.

Mica shares the methods behind Augury’s fault testing processes, why they use the highest quality data available, how in-house experts help them filter their data reliably, and their approach to communicating with customers. Our conversation also explores the balance between edge computing and cloud computing, and why both are necessary for optimal performance and monitoring.

Key Points From This Episode:

• Mica’s journey from studying physics at the Weizmann Institute to her current role at Augury.
• How her background in physics and neuroscience inform her work in AI.
• Why physicists are drawn to AI and data science; how scientists follow their curiosity.
• Mica’s responsibilities in her role as algorithms team lead at Augury.
• How they develop algorithms and test for faults; why this requires the highest quality data.
• Understanding the role of their in-house expert vibration analysts.
• The importance of domain expertise in labeling and annotating data.
• Finding the balance between manual and automated processes in data labeling.
• How to communicate with customers and present metrics that matter to them.
• Augury’s use of edge and cloud computing for optimal performance and monitoring.

Quotes:

“We look for better ways to adjust our algorithms and also develop new ones for all kinds of faults that could happen in the machines catching events that are trickier to catch, and for that we need highest quality data.” — Mica Rubinson [0:08:20]

“At Aubrey, we have internal vibration analysts that are experts in their field. They go through very rigorous training process. There are international standards to how you do vibration analysis, and we have them in-house.” — Mica Rubinson [0:09:07]

“[It’s] really helpful for us to have [these] in-house experts. We have massive amounts of records – signal recordings from 10 years of machine monitoring. Thanks to these experts [in] labeling, we can filter out a lot of noisy parts of this data.” — Mica Rubinson [0:10:32]

“We quantify [our services] for the customer as their ROI [and] how much they saved by using Augury. You had this [issue, and] we avoided this downtime. [We show] how much does it translates eventually [into] money that you saved.” — Mica Rubinson [0:22:28]

Links Mentioned in Today’s Episode:

Mica Rubinson on LinkedIn

Mica Rubinson on ResearchGate

Augury
Weizmann Institute of Science

How AI Happens

Sama

Srini highlights the importance of integrating these agents into real-world applications, enhancing productivity and user experiences across industries. Srini also delves into the challenges of building reliable, ethical, and secure AI systems while fostering developer innovation. His insights offer a roadmap for harnessing advanced agents to drive meaningful technological progress. Don’t miss this informative conversation. 

Key Points From This Episode:

• Introducing today’s guest, Srini Iragavarapu, a leader at AWS. 
• His thoughts on how Agentic and AI are intersecting today. 
• The state of the union of agents in the world and at AWS.
• How AWS is leveraging agents to build specific tasks for customers.
• Two mechanisms that software agents use to operate.
• Understanding the reasoning capabilities of large foundational models.
• How AWS makes use of a test agent. 
• Qdeveloper’s instantaneous conversational capabilities.
• Bringing different options to the customers as a long-term strategy.
• Three layers at which AWS is innovating today.
• Why the end user is ultimately the person who benefits. 

Quotes:

“Think of it as an iterative way of solving a problem rather than just calling a single API and coming back: that’s in a nutshell how generative AI and the foundation models are working with reasoning capabilities.” — Srini Iragavarapu [0:03:04]

“The models are becoming more powerful and more available, faster,  a lot more dependable.” — Srini Iragavarapu [0:29:57]

Links Mentioned in Today’s Episode:

Srini Iragavarapu on LinkedIn

How AI Happens

Sama

 We explore the current trends of AI-based solutions in retail, what has driven its adoption in the industry, and how AI-based customer service technology has improved over time. We also discuss the correct mix of technology and humans, the importance of establishing boundaries for AI, and why it won't replace humans but will augment workflow. Hear examples of AI retail success stories, what companies got AI wrong, and the reasons behind the wins and failures. Gain insights into the value of copilots, business strategies to avoid investing in ineffective AI solutions, and much more. Tune in now!

Key Points From This Episode:

• Learn about Lisa and Mika's backgrounds in retail technology and AI-based solutions. 
• Hear how AI has become more accessible to businesses beyond the typical tech giants.
• Explore how AI-powered chatbots and copilots have evolved to improve customer service.
• The Coca-Cola AI ad controversy and why oversight on AI-generated content is vital.
• Discover the innovative and exciting ways AI can be leveraged in the retail industry.
• AI success stories: Target’s AI copilot for employees and Nordstrom’s personalization tool.
• How AI is making the return process more efficient and improving inventory management.
• Uncover the multimodal connections of AI and how it will enhance customer personalization.
• Important considerations for businesses regarding the adoption of AI and the pitfalls to avoid.

Quotes:

“I think [the evolution] in terms of accessibility to AI-solutions for people who don't have the massive IT departments and massive data analytics departments is really remarkable.” — Mika Yamamoto [0:04:25]

“Whether it's generative AI for creative or content or whatever, it's not going to replace humans. It's going to augment our workflows.” — Lisa Avvocato [0:10:46]

“Retail is actually one of the fastest adopting industries out there [of] AI.” — Mika Yamamoto [0:14:17]

“Having conversations with peers, I think, is absolutely invaluable to figure out what's hype and what's reality [regarding AI].” — Mika Yamamoto [0:30:19]

Links Mentioned in Today’s Episode:

Lisa Avvocato on LinkedIn

Mika Yamamoto on LinkedIn

Freshworks

The Coca‑Cola Company

How AI Happens

Sama

We hear about Nitzan’s AI expertise, motivation for joining eBay, and approach to implementing AI into eBay's business model. Gain insights into the impacts of centralizing and federating AI, leveraging generative AI to create personalized content, and why patience is essential to AI development. We also unpack eBay's approach to LLM development, tailoring AI tools for eBay sellers, the pitfalls of generic marketing content, and the future of AI in retail. Join us to discover how AI is revolutionizing e-commerce and disrupting the retail sector with Nitzan Mekel-Bobrov! 

Key Points From This Episode:

• Nitzan's career experience, his interest in sustainability, and his sneaker collection.
• Why he decided to begin a career at eBay and his role at the company.
• His approach to aligning the implementation of AI with eBay's overall strategy.
• How he identifies the components of eBay's business model that will benefit from AI.
• What makes eBay highly suitable for the implementation of AI tools.
• Challenges of using generative AI models to create personalized content for users.
• Why experimentation is vital to the AI development and implementation process.
• Aspects of the user experience that Nitzan uses to train and develop eBay's LLMs.
• The potential of knowledge graphs to uncover the complexity of user behavior.
• Reasons that the unstructured nature of eBay's data is fundamental to its business model.
• Incorporating a seller's style into AI tools to avoid creating generic marketing material.
• Details about Nitzan’s team and their diverse array of expertise.
• Final takeaways and how companies can ensure they survive the AI transition. 

Quotes:

“It’s tricky to balance the short-term wins with the long-term transformation.” — Nitzan Mekel-Bobrov [0:06:50]

“An experiment is only a failure if you haven’t learned anything yourself and – generated institutional knowledge from it.” — Nitzan Mekel-Bobrov [0:09:36]

“What's nice about [eBay's] business model — is that our incentive is to enable each seller to maintain their own uniqueness.” — Nitzan Mekel-Bobrov [0:27:33]

“The companies that will thrive in this AI transformation are the ones that can figure out how to marry parts of their current culture and what all of their talent brings with what the AI delivers.” — Nitzan Mekel-Bobrov [0:33:58]

Links Mentioned in Today’s Episode:

Nitzan Mekel-Bobrov on LinkedIn

eBay

How AI Happens

Sama

Satya unpacks how Unilever utilizes its database to inform its models and how to determine the right amount of data needed to solve complex problems. Dr. Wattamwar explains why contextual problem-solving is vital, the notion of time constraints in data science, the system point of view of modeling, and how Unilever incorporates AI into its models. Gain insights into how AI can increase operational efficiency, exciting trends in the AI space, how AI makes experimentation accessible, and more! Tune in to learn about the power of data science and AI with Dr. Satyajit Wattamwar.

Key Points From This Episode:

• Background on Dr. Wattamwar, his PhD research, and data science expertise.
• Unpacking some of the commonalities between data science and physics. 
• Why the outcome of using significantly large data sets depends on the situation. 
• The minimum amount of data needed to make meaningful and quality models.
• Examples of the common mistakes and pitfalls that data scientists make.
• How Unilever works with partner organizations to integrate AI into its models.
• Ways that Dr. Wattamwar uses AI-based tools to increase his productivity.
• The difference between using AI for innovation versus operational efficiency.
• Insight into the shifting data science landscape and advice for budding data scientists.

Quotes:

“Around – 30 or 40 years ago, people started realizing the importance of data-driven modeling because you can never capture physics perfectly in an equation.” — Dr. Satyajit Wattamwar [0:03:10]

“Having large volumes of data which are less related with each other is a different thing than a large volume of data for one problem.” — Dr. Satyajit Wattamwar [0:09:12]

“More data [does] not always lead to good quality models. Unless it is for the same use-case.” — Dr. Satyajit Wattamwar [0:11:56]

“If somebody is looking [to] grow in their career ladder, then it's not about one's own interest.” — Dr. Satyajit Wattamwar [0:24:07]

Links Mentioned in Today’s Episode:

Dr. Satyajit Wattamwar on LinkedIn

Unilever

How AI Happens

Sama

Jing explains how Vanguard uses machine learning and reinforcement learning to deliver personalized "nudges," helping investors make smarter financial decisions. Jing dives into the importance of aligning AI efforts with Vanguard’s mission and discusses generative AI’s potential for boosting employee productivity while improving customer experiences. She also reveals how generative AI is poised to play a key role in transforming the company's future, all while maintaining strict data privacy standards.

Key Points From This Episode:

• Jing Wang’s time at Fermilab and the research behind her PhD in high-energy physics.
• What she misses most about academia and what led to her current role at Vanguard.
• How she aligns her team’s AI strategy with Vanguard’s business goals.
• Ways they are utilizing AI for nudging investors to make better decisions.
• Their process for delivering highly personalized recommendations for any given investor.
• Steps that ensure they adhere to finance industry regulations with their AI tools.
• The role of reinforcement learning and their ‘next best action’ models in personalization.
• Their approach to determining the best use of their datasets while protecting privacy.
• Vanguard’s plans for generative AI, from internal productivity to serving clients.
• How Jing stays abreast of all the latest developments in physics.

Quotes:

“We make sure all our AI work is aligned with [Vanguard’s] four pillars to deliver business impact.” — Jing Wang [0:08:56]

“We found those simple nudges have tremendous power in terms of guiding the investors to adopt the right things. And this year, we started to use a machine learning model to actually personalize those nudges.” — Jing Wang [0:19:39]

“Ultimately, we see that generative AI could help us to build more differentiated products. – We want to have AI be able to train language models [to have] much more of a Vanguard mindset.” — Jing Wang [0:29:22]

Links Mentioned in Today’s Episode:

Jing Wang on LinkedIn

Vanguard
Fermilab
How AI Happens

Sama

Key Points From This Episode:

• Ram Venkatesh describes his career journey to founding Sema4.ai. 
• The pain points he was trying to ease with Sema4.ai.
• How our general approach to big data is becoming more streamlined, albeit rather slowly. 
• The ins and outs of Sema4.ai and how it serves its clients. 
• What Ram means by “agent” and “agent agency” when referring to machine learning copilots.
• The difference between writing a program to execute versus an agent reasoning with it.  
• Understanding the contextual work training method for agents. 
• The relationship between an LLM and an agent and the risks of training LLMs on agent data.
• Exploring the next generation of LLM training protocols in the hopes of improving efficiency. 
• The requirements of an LLM if you’re not training it and unpacking modality improvements. 
• Why agent input and feedback are major disruptions to SaaS and beyond. 
• Our guest shares his hopes for the future of AI. 

Quotes:

“I’ve spent the last 30 years in data. So, if there’s a database out there, whether it’s relational or object or XML or JSON, I’ve done something unspeakable to it at some point.” — @ramvzz [0:01:46]

“As people are getting more experienced with how they could apply GenAI to solve their problems, then they’re realizing that they do need to organize their data and that data is really important.” — @ramvzz [0:18:58]

“Following the technology and where it can go, there’s a lot of fun to be had with that.” — @ramvzz [0:23:29]

“Now that we can see how software development itself is evolving, I think that 12-year-old me would’ve built so many more cooler things than I did with all the tech that’s out here now.” — @ramvzz [0:29:14]

Links Mentioned in Today’s Episode:

Ram Venkatesh on LinkedIn

Ram Venkatesh on X

Sema4.ai

Cloudera

How AI Happens

Sama

Pascal & Yannick delve into the kind of human involvement SAM-2 needs before discussing the use cases it enables. Hear all about the importance of having realistic expectations of AI, what the cost of SAM-2 looks like, and the the importance of humans in LLMs.

Key Points From This Episode:

• Introducing Pascal Jauffret and Yannick Donnelly to the show.
• Our guests explain what the SAM-2 model is. 
• A description of what getting information from video entails.
• What made our guests interested in researching SAM-2. 
• A few things that stand out about this tool. 
• The level of human involvement that SAM-2 needs. 
• Some of the use cases they see SAM-2 enabling. 
• Whether manually annotating is easier than simply validating data. 
• The importance of setting realistic expectations of what AI can do. 
• When LLM models work best, according to our experts.
• A discussion about the cost of the models at the moment. 
• Why humans are so important in coaching people to use models. 
• What we can expect from Sama in the near future. 

Quotes:

“We’re kind of shifting towards more of a validation period than just annotating from scratch.” — Yannick Donnelly [0:22:01]

“Models have their place but they need to be evaluated.” — Yannick Donnelly [0:25:16]

“You’re never just using a model for the sake of using a model. You’re trying to solve something and you’re trying to improve a business metric.” — Pascal Jauffret [0:32:59]

“We really shouldn’t underestimate the human aspect of using models.” — Pascal Jauffret [0:40:08]

Links Mentioned in Today’s Episode:

Pascal Jauffret on LinkedIn

Yannick Donnelly on LinkedIn

How AI Happens

Sama

Today we are joined by Siddhika Nevrekar, an experienced product leader passionate about solving complex problems in ML by bringing people and products together in an environment of trust.  We  unpack the state of free computing, the challenges of training AI models for edge, what Siddhika hopes to achieve in her role at Qualcomm, and her methods for solving common industry problems that developers face.

Key Points From This Episode:

• Siddhika Nevrekar walks us through her career pivot from cloud to edge computing. 
• Why she’s passionate about overcoming her fears and achieving the impossible. 
• Increasing compute on edge devices versus developing more efficient AI models.
• Siddhika explains what makes Apple a truly unique company. 
• The original inspirations for edge computing and how the conversation has evolved. 
• Unpacking the current state of free computing and what may happen in the near future. 
• The challenges of training AI models for edge. 
• Exploring Siddhika’s role at Qualcomm and what she hopes to achieve. 
• Diving deeper into her process for achieving her goals. 
• Common industry challenges that developers are facing and her methods for solving them

Quotes:

“Ultimately, we are constrained with the size of the device. It’s all physics. How much can you compress a small little chip to do what hundreds and thousands of chips can do which you can stack up in a cloud? Can you actually replicate that experience on the device?” — @siddhika_ 

“By the time I left Apple, we had 1000-plus [AI] models running on devices and 10,000 applications that were powered by AI on the device, exclusively on the device. Which means the model is entirely on the device and is not going into the cloud. To me, that was the realization that now the moment has arrived where something magical is going to start happening with AI and ML.” — @siddhika_ 

Links Mentioned in Today’s Episode:

Siddhika Nevrekar on LinkedIn

Siddhika Nevrekar on X

Qualcomm AI Hub

How AI Happens

Sama