Transformation of the Week
Angela Chapman

Transcript

Lusk: Hey, Angela, thanks for being here. Congratulations on winning the Transformation Of The Week. Tell me a little bit more about your background and what you’re doing today.

Angela: Thanks, Chris. I’m glad to be here. I have over a decade of experience in AI and data science, starting a long time ago with a master’s degree in cognitive and Neural systems, where we applied machine learning to study the human brain. Since then, I’ve held both hands-on and leadership roles in AI. So that includes being a former director of data science at Johnson and Johnson, where my team focused on deep learning for digital surgery.

The past four years, I’ve been running my own consulting practice, which is Blue Grotto Labs, LLC. And I build AI prototypes, help companies get AI products into production from a prototype stage, and advise on AI strategy across many industries. So over the past two years, as you can imagine, that’s been almost all generative AI implementations and prototypes of different kinds. It’s been really exciting to see the wide array of things that people are doing with it.

Lusk: That’s a very impressive background. What got you interested, specifically, in gen AI?

Angela: Well, over the past two years, it’s been impossible not to work in generative AI, if you’re running an AI consulting business. This is what a lot of companies are really excited about right now. It excites me because it’s been such a big paradigm shift in AI development. You can so easily build capabilities that were really hard to do five years ago, and there’s always new stuff to learn. So I really enjoy the space.

I’ve worked on more than ten generative AI projects in the past year in industries as diverse as health care, e-commerce, HR, even an HVAC technology firm. And the biggest challenge is always moving from a cool prototype to a really robust production system. I think oftentimes I’ve seen clients underestimate the effort, and cost, and time it takes to build something that’s really production ready in the space.

Lusk: So you graduated from our AI Engineering Bootcamp number four, just about two months ago. Have you been able to put anything that you’ve learned into practice, into your consultancy?

Angela: Yeah, absolutely. Although I had experience in LLMs before the course, I took it to explore cutting edge open source tools and, you know, learn some approaches and things that I don’t get to do in my day to day work.

And since then, I’ve used a lot of the tools that we used in the course, in my own consulting work, including RAGAS, which has been great for evaluation and synthetic data generation. I’ve been using LangSmith quite a bit for tracing and monitoring, and Qdrant for vector search.

I’ve also put into practice, not just tools, but design patterns that I learned in the course. For example, I’ve started using agents in a lot more workflows that a year ago would have been vanilla RAG systems. I think that there are a lot of use cases for agents, even though multi-agent systems are still early stage as far as production goes. So I think a lot of what I’ve learned in the course has been really helpful in my consulting practice.

Lusk: So what’s a big AI project that you’re currently working on?

Angela: Well, a big thing that I’m working on right now, has to do with my AI Makerspace project. I developed an agentic RAG chatbot to support family caregivers of dementia patients. You know, I presented it at the AI makerspace demo day at the end of the bootcamp, as well as at the MLOps World Conference in Austin recently.

And although I was thinking of it as more of just a demo project, it got a lot of positive interest from both academic and industry stakeholders. So I’m taking it further. I’m currently applying for NIH funding to take this development to the next level with an academic partner. And I’m really excited about this project and couldn’t have done it without the AI Makerspace Bootcamp.

Lusk: So what words of wisdom do you have for someone out there that’s watching this who’s considering moving into the Gen AI space?

Angela: I would say generative AI is a really exciting industry to be in right now, so if you are interested in going for it, just start building stuff, start learning, I really recommend it.

Couple of learnings that I’ve had over my past couple years of working in the space. First, learn core machine learning principles or bring them along with you if you already have them. It’s easy to string together APIs and create a cool prototype in generative AI, but strong data science fundamentals like experimentation practices and train test, validation and understanding model architectures is still just as important as it was five years ago.

Another big learning, I would say, try to understand production challenges and the gap between a prototype and production. Start thinking early about things like latency, scaling, cost, guardrails, and how to really make things work at scale and that’s going to set you apart in your machine learning work. But again, it’s a great place to be right now. Very, very exciting. Always new stuff to learn. And I encourage anyone who’s thinking about it to dive in.

Lusk: Angela, you’re building, shipping, and sharing like a legend, it’s the reason you won this week’s Transformation Of The Week. Tell the people out there how they can connect with you.

Angela: Well, you can find me on LinkedIn. You can find me on GitHub. And you could also go to www.BlueGrottoLabs.com and contact me there. I’m always excited to meet others who are working in the industry or considering it.

So if you’re a company that would like to stand up a prototype or looking for help getting a prototype into production in the generative AI space, I’m happy to connect with you and talk through what that might look like as well.