The MLOps Podcast

Dean Pleban and Liron Itzhakhi Allerhand explore what it really takes to move LLMs into production. They cover how to define clear requirements, prep data for RAG, engineer effective prompts, and evaluate model performance using concrete metrics. The conversation dives into managing sensitive data, avoiding leakage, and why crisp outputs and clear user intent matter. Plus: future trends like in-context learning and the decoupling of foundation models from vertical apps.
Join our Discord community:
https://discord.gg/tEYvqxwhah
---
Timestamps:
00:00 Introduction
01:48 Phases of LLM Project Development
03:32 Defining the Problem
09:35 Data Preparation and Understanding
23:59 Multimodal RAG
26:28 Prompt Engineering & Model Selection
27:58 Model Fine-tuning & Customization
33:18 LLM as a Judge
38:58 Evaluating Model Performance and Handling Hallucinations
41:02 Using LLMs with sensitive data
45:24 Other ideas for model evaluation and guardrails
49:28 Recommendations for the audience

➡️ Liron Itzhaki Allerhand on LinkedIn – https://www.linkedin.com/in/liron-izhaki-allerhand-16579b4/
🌐 Check Out Our Website! https://dagshub.com Social
Links:
➡️ LinkedIn: https://www.linkedin.com/company/dagshub
➡️ Twitter: https://x.com/TheRealDAGsHub
➡️ Dean Pleban: https://x.com/DeanPlbn

In this episode, Dean and Natanel Davidovits explore the intricacies of AI and machine learning, focusing on model efficiency, the use of APIs versus self-hosting, and the importance of defining success metrics in real-world applications. They discuss the challenges of data quality and labeling, the evolving role of data scientists in the age of LLMs, and the significance of effective communication between data science and product teams. The conversation also touches on the future of robotics in AI and the need for specialization in a rapidly changing landscape.

Join our Discord community: https://discord.gg/tEYvqxwhah

---
Timestamps:
00:00 Introduction to Natanel Davidovits
02:10 Optimizing AI Models for Real-World Tasks
03:47 Success Metrics in Industry vs. Academia
07:52 The Importance of Communication Between Teams
11:33 Handling Data Quality and Labeling Challenges
12:11 The Impact of LLMs on Data Science Careers
16:29 Navigating Specialized Domain Data
22:15 Trends in Machine Learning and AI
27:27 The Future of AI and Robotics
28:28 The Role of AI in Physics
33:36 Controversial Views on AI and Machine Learning
34:05 Final Thoughts and Recommendations

➡️ Natanel Davidovits on LinkedIn – https://www.linkedin.com/in/natanel-davidovits-28695312/

🌐 Check Out Our Website! https://dagshub.com

Social Links:
➡️ LinkedIn: https://www.linkedin.com/company/dagshub
➡️ Twitter: https://x.com/TheRealDAGsHub
➡️ Dean Pleban: https://x.com/DeanPlbn

In this episode, Dean speaks with Jeremie Dreyfuss, Head of AI Research and Development at Intel, about the evolving role of AI in the enterprise. Jeremie shares insights into scaling machine learning solutions, the challenges of building AI infrastructure, and the future of AI-driven innovation in large organizations. Learn how enterprises are leveraging AI for efficiency, the latest advancements in AI research, and the strategies for staying competitive in a rapidly changing landscape.

Join our Discord community: https://discord.gg/tEYvqxwhah

---
Timestamps:
00:00 Introduction and Overview
00:55 Challenges of Data Collection and Infrastructure
05:00 Optimizing Test Recommendations
14:42 Tips for Deploying Entire ML Pipelines
21:19 The Impact of Large Language Models (LLMs)
25:30 How to Decide About LLM Investment in the Enterprise
29:29 Evaluating Models and Using Synthetic Data
35:34 Choosing the Right Tools for ML and LLM Projects
45:21 The Beauty of Small Data in Machine Learning
48:22 Recommendations for the Audience

➡️ Jeremie Dreyfuss on LinkedIn – https://www.linkedin.com/in/jeremie-dreyfuss/

🌐 Check Out Our Website! https://dagshub.com

Social Links:
➡️ LinkedIn: https://www.linkedin.com/company/dagshub
➡️ Twitter: https://x.com/TheRealDAGsHub
➡️ Dean Pleban: https://x.com/DeanPlbn

In this episode, Dean speaks with Dror Haor, CTO at SeeTree, about the challenges of deploying AI in agriculture at scale. They explore how SeeTree integrates AI and sensor fusion to manage vast amounts of remote sensing data, helping farmers improve crop yields with high accuracy at low costs. Dror shares insights on handling data drift, customizing models for different regions, and balancing the trade-offs between cost and performance. This conversation dives deep into practical machine learning applications in agriculture, offering valuable lessons for anyone working with large-scale data and AI.

Join our Discord community: https://discord.gg/tEYvqxwhah

---
Timestamps:
00:00 Introduction
00:32 Production in machine learning at SeeTree
07:34 Sensor fusion in machine learning
16:26 Balancing accuracy and cost in agriculture
20:09 Customizing models for different customers and crops
24:19 Dealing with data in different domains
30:10 Tools and processes for ML at SeeTree
35:58 Building for scale
40:17 Collecting user feedback and self-improving products
42:45 Exciting developments in ML & AI
45:12 Hot takes in ML - Overfitting is good
46:34 Recommendations for the Audience

➡️ Dror Haor on LinkedIn – https://www.linkedin.com/in/dror-haor-phd-77152322/
➡️ Dror Haor on Twitter – https://x.com/DrorHaor

🌐 Check Out Our Website! https://dagshub.com

Social Links:
➡️ LinkedIn: https://www.linkedin.com/company/dagshub
➡️ Twitter: https://x.com/TheRealDAGsHub
➡️ Dean Pleban: https://x.com/DeanPlbn

In this episode, Dean speaks with Federico Bacci, a data scientist and ML engineer at Bol, the largest e-commerce company in the Netherlands and Belgium. Federico shares valuable insights into the intricacies of deploying machine learning models in production, particularly for forecasting problems. He discusses the challenges of model explainability, the importance of feature engineering over model complexity, and the critical role of stakeholder feedback in improving ML systems. Federico also offers a compelling perspective on why LLMs aren't always the answer in AI applications, emphasizing the need for tailored solutions. This conversation provides a wealth of practical knowledge for data scientists and ML engineers looking to enhance their understanding of real-world ML operations and challenges in e-commerce.

Join our Discord community: https://discord.gg/tEYvqxwhah

---
Timestamps:
00:00 Introduction and Background
01:59 Owning the ML Pipeline
02:56 Deployment Process
05:58 Testing and Feedback
07:40 Different Deployment Strategies
11:19 Explainability and Feature Importance
13:46 Challenges in Forecasting
22:33 ML Stack and Tools
26:47 Orchestrating Data Pipelines with Airflow
31:27 Exciting Developments in ML
35:58 Recommendations and Closing

Links
Dwarkesh podcast with Anthropic and Gemini team members – https://www.dwarkeshpatel.com/p/sholto-douglas-trenton-bricken

➡️ Federico Bacci on LinkedIn – https://www.linkedin.com/in/federico-bacci/
➡️ Federico Bacci on Twitter – https://x.com/fedebyes

🌐 Check Out Our Website! https://dagshub.com

Social Links:
➡️ LinkedIn: https://www.linkedin.com/company/dagshub
➡️ Twitter: https://x.com/TheRealDAGsHub
➡️ Dean Pleban: https://x.com/DeanPlbn