Deploying LLMs with Two Data-Savvy Dads: Chris Brousseau and Matthew Sharp

Welcome back to episode #29 of the FITFO podcast! Today, we have a fascinating episode featuring AI experts Chris Brousseau and Matthew Sharp. They join us to discuss their new book, "LLMs in Production," which aims to help beginners simplify complex machine learning concepts and guide practitioners in deploying reliable, real-world AI models.

In this episode, we'll explore their transition from data science to engineering, the differences between data scientists and machine learning engineers, and the significance of reinforcement learning with human feedback (RLHF). Chris and Matthew also delve into critical topics like tokenization, embeddings, and the strategic decision-making involved in building versus buying AI solutions.

They also share insights on hiring the right talent when starting out your own AI/ML team and they drop other nuggets of knowledge for anyone interested in the practical aspects of AI and machine learning. Tune in for an informative discussion that’s set to expand your understanding and skills in the ever-evolving world of AI!

Meet the Guests:

Matt Sharp is an engineer and seasoned technology leader in MLOps. Has led successful initiatives for both startups and top-tier tech companies. Matt specializes in deploying, managing, and scaling machine learning models. ⁠https://www.linkedin.com/in/matthewsharp/⁠ 

Christopher Brousseau is a Principal Machine Learning Engineer with a linguistics and localization background. He specializes in linguistically-informed NLP, especially with an international focus and has led successful ML and Data product initiatives at both startups and Fortune 500s. https://www.linkedin.com/in/chris-brousseau/⁠

Links From Show:

LLMs in Production: https://mng.bz/JZlZ Use Coupon Code FITFO24 for 30% off

Github Repository: https://github.com/IMJONEZZ/LLMs-in-Production

Link to Forge Salt Lake City MeetUp: https://www.meetup.com/pro/forge-utah/

Time Stamps:

00:00 Podcast episode features three dads discussing AI.

09:19 Writing for different platforms requires focused content.

12:12 Demystifying technical content for beginners and beyond.

15:46 About their upcoming Book LLMS in Production

23:57 ML Engineering Interview tip

28:46 Simplifying machine learning through predicting data lines.

34:48 Building reliable, user-friendly products with impactful data.

47:14 Teaching robot to do backflip through feedback.

57:52 Understanding language concepts vital for machine learning.

01:07:33 the significance of the transformer model.

01:11:25 Tokenization strategy affects model performance and significance.

01:18:53 Evaluating AI/ML Talent

01:24:28 Understanding the time taken for model generation.

01:25:40 To process 86,400 things in a day.