Log in

goodpods headphones icon

To access all our features

Open the Goodpods app
Close icon
Data Radicals - LLMs Decoded: A Starter's Guide to AI with Raza Habib, co-founder & CEO of Humanloop
plus icon
bookmark

LLMs Decoded: A Starter's Guide to AI with Raza Habib, co-founder & CEO of Humanloop

02/05/25 • 48 min

Data Radicals

As AI becomes integral to every aspect of business, ensuring its accessibility for everyone—not just specialists—is essential. Companies like Humanloop are leading the charge with innovative platforms that empower non-technical users to harness the power of advanced language models through intuitive tools and frameworks.

Democratizing AI access paves the way for transformative business outcomes and a future of collaborative AI systems. However, building a strong AI strategy starts with leveraging powerful models and mastering prompt engineering before considering fine-tuning. Engaging subject matter experts and using robust evaluation and collaboration tools are equally critical to the success of modern AI projects. In this episode, Satyen and Raza examine the evolution of AI models, the practical challenges of model evaluation and prompt engineering, and the role of multidisciplinary teams in AI development.

*Satyen’s narration was created using AI

--------

“ In our experience, fine-tuning is very useful as an optimization step. But, it's not where we recommend people to start. When people are trying to customize these models, we encourage them as much as possible to push the limits of prompt engineering with the most powerful model they can before they consider fine-tuning. The reason that we suggest that is that it's much faster to change a prompt and see what the impact is. It's often sufficient to customize the models and it's less destructive. If you fine-tune a model and you want to update it later, you kind of have to start from scratch. You have to go back to the base model with your label data set and re fine-tune from the beginning. If you're customizing the model via prompts and you want to make a change, you just go change the text and you can see the difference. There's a much faster iteration cycle and you can get most of the benefit.” – Raza Habib

--------

Time Stamps

*(01:26): Raza’s career journey: From academia to industry

*(12:46): What is active learning?

*(17:20): How LLMs diverge from traditional software processes

*(24:53): What is data leakage?

*(35:56): How can software engineers adapt in the age of AI?

*(47:04): Satyen’s takeaways

--------

Sponsor

This podcast is presented by Alation.

Learn more:

Subscribe to the newsletter: https://www.alation.com/podcast/

Alation’s LinkedIn Profile: https://www.linkedin.com/company/alation/

Satyen’s LinkedIn Profile:

https://www.linkedin.com/in/ssangani/

--------

Links

Connect with Raza on LinkedIn

Learn more about Humanloop

Listen to Raza’s podcast

Order Information Theory, Inference and Learning Algorithms by David MacKay

plus icon
bookmark

As AI becomes integral to every aspect of business, ensuring its accessibility for everyone—not just specialists—is essential. Companies like Humanloop are leading the charge with innovative platforms that empower non-technical users to harness the power of advanced language models through intuitive tools and frameworks.

Democratizing AI access paves the way for transformative business outcomes and a future of collaborative AI systems. However, building a strong AI strategy starts with leveraging powerful models and mastering prompt engineering before considering fine-tuning. Engaging subject matter experts and using robust evaluation and collaboration tools are equally critical to the success of modern AI projects. In this episode, Satyen and Raza examine the evolution of AI models, the practical challenges of model evaluation and prompt engineering, and the role of multidisciplinary teams in AI development.

*Satyen’s narration was created using AI

--------

“ In our experience, fine-tuning is very useful as an optimization step. But, it's not where we recommend people to start. When people are trying to customize these models, we encourage them as much as possible to push the limits of prompt engineering with the most powerful model they can before they consider fine-tuning. The reason that we suggest that is that it's much faster to change a prompt and see what the impact is. It's often sufficient to customize the models and it's less destructive. If you fine-tune a model and you want to update it later, you kind of have to start from scratch. You have to go back to the base model with your label data set and re fine-tune from the beginning. If you're customizing the model via prompts and you want to make a change, you just go change the text and you can see the difference. There's a much faster iteration cycle and you can get most of the benefit.” – Raza Habib

--------

Time Stamps

*(01:26): Raza’s career journey: From academia to industry

*(12:46): What is active learning?

*(17:20): How LLMs diverge from traditional software processes

*(24:53): What is data leakage?

*(35:56): How can software engineers adapt in the age of AI?

*(47:04): Satyen’s takeaways

--------

Sponsor

This podcast is presented by Alation.

Learn more:

Subscribe to the newsletter: https://www.alation.com/podcast/

Alation’s LinkedIn Profile: https://www.linkedin.com/company/alation/

Satyen’s LinkedIn Profile:

https://www.linkedin.com/in/ssangani/

--------

Links

Connect with Raza on LinkedIn

Learn more about Humanloop

Listen to Raza’s podcast

Order Information Theory, Inference and Learning Algorithms by David MacKay

Previous Episode

undefined - Empowering AI Practitioners with Wendy Turner-Williams, CEO of TheAssociation.AI

Empowering AI Practitioners with Wendy Turner-Williams, CEO of TheAssociation.AI

There’s a digital revolution happening – and it’s poised to impact data leaders across all industries. During this time of never-ending change, it’s crucial to have data practitioners at the center of holistic AI transformation as regulatory compliance and ethical standards come into the fold.

Businesses of every size will encounter these complex regulations. Learn about these challenges and how connecting practitioners across fields can create more compliant and trusted AI environments.

This episode is packed with practical guidelines and future-focused strategies designed to empower data leaders with the insights they need to build effective, ethical AI.

*Satyen’s narration was created using AI

--------

“ Each state or each country having their own AI policies or privacy policies, frankly, doesn't make any sense. Because, most people, especially if you're on cloud, you may not even know where your data sits. There's basic principles and there's basic practices that you can define that are tech agnostic, that you can still have your own tech stack and your own tools. There's lots of solutions and players that work in those components, but you can give basic guidelines to say, here's the steps and the processes and the pieces that you need to put in place. Here's how they form together to create an encapsulation of trust.” – Wendy Turner-Williams

--------

Time Stamps

*(02:20): Enabling the AI community

*(13:49): How does The Association.AI put regulatory theory into practice?

*(21:01): Why AI practitioners need places to knowledge share

*(34:23): The rise of the CIO: Risk talks

*(39:46): AI predictions: What will change? What won’t?

*(50:30): Satyen’s takeaways

--------

Sponsor

This podcast is presented by Alation.

Learn more:

Subscribe to the newsletter: https://www.alation.com/podcast/

Alation’s LinkedIn Profile: https://www.linkedin.com/company/alation/

Satyen’s LinkedIn Profile:

https://www.linkedin.com/in/ssangani/

--------

Links

Connect with Wendy on LinkedIn

Learn more about TheAssociation.AI

Order Unleashing the Power of Data with Trusted AI

Next Episode

undefined - Redesigning Processes for the Age of AI Agents with Tom Davenport

Redesigning Processes for the Age of AI Agents with Tom Davenport

AI is here—but are businesses truly ready to harness its full potential? In this episode of Data Radicals, host Satyen Sangani sits down with Tom Davenport to explore what it takes for AI to create real business value.

As one of the most respected voices in AI and analytics, Tom brings decades of expertise to the table. From agentic AI to AI-driven leadership, this conversation covers the pressing challenges—and opportunities—that will shape the next era of business transformation.

What You'll Learn in This Episode:

🔹 Agentic AI is still in its infancy – AI agents can act autonomously, but most businesses are only using them for simple, low-stakes tasks. Scaling their use requires overcoming reliability challenges.
🔹 AI alone won’t drive value—process redesign is critical – Businesses can’t just add AI to existing workflows and expect results. As Tom puts it, "Economic value requires that we change the way we do our work. And there has to be some intentional design activity. It can't just evolve."
🔹 The C-suite is overcrowded—AI leadership must evolve – With CIOs, CDOs, CTOs, and more, organizations often suffer from fragmented leadership. Tom argues for business-driven executives who can oversee AI, data, and digital strategy holistically.

AI is transforming industries, but the organizations that truly succeed will be those that rethink their leadership, workflows, and data strategies. Whether you're a CDO, CIO, or data professional, this episode offers actionable insights from one of the most influential thinkers in AI and analytics.

*Satyen’s narration was created using AI

--------

“ There is a generative AI component, but it's not just generative AI. It's probably analytical AI as well, it's probably still APIs, it's probably still transaction systems, ERP and CRM and so on. There'll have to be a lot of integration, which means that it's going to be a fair amount of work for companies to pull this off. I think vendors will help and they'll provide lots of tools, but I think companies will have to figure out what they want to accomplish with it and make it happen and that will take some time and effort.” – Tom Davenport

--------

Time Stamps

*(02:27): Agentic AI use cases: Is AI the new software?

*(10:52): The CDO's role in the age of AI

*(20:57): What is AGI? Is it coming soon?

*(26:43): How can organizations transform with data?

*(35:38): The need to redesign processes for AI

*(38:58): Satyen’s takeaways

--------

Sponsor

This podcast is presented by Alation.

Learn more:

Subscribe to the newsletter: https://www.alation.com/podcast/

Alation’s LinkedIn Profile: https://www.linkedin.com/company/alation/

Satyen’s LinkedIn Profile:

https://www.linkedin.com/in/ssangani/

--------

Links

Connect with Tom on LinkedIn

Read Tom’s MIT Sloan article Five Trends in AI and Data Science for 2025

Read Tom’s Harvard Business Review article How Gen AI and Analytical AI Differ — and When to Use Each

Order Tom’s book All Hands on Tech: The AI-Powered Citizen Revolution

Order Tom’s book All in on AI: How Smart Companies Win Big with Artificial Intelligence

Order Tom’s book Competing on Analytics: The New Science of Winning

Episode Comments

Generate a badge

Get a badge for your website that links back to this episode

Select type & size
Open dropdown icon
share badge image

<a href="https://goodpods.com/podcasts/data-radicals-217063/llms-decoded-a-starters-guide-to-ai-with-raza-habib-co-founder-and-ceo-83631515"> <img src="https://storage.googleapis.com/goodpods-images-bucket/badges/generic-badge-1.svg" alt="listen to llms decoded: a starter's guide to ai with raza habib, co-founder & ceo of humanloop on goodpods" style="width: 225px" /> </a>

Copy