TestGuild Automation Podcast Testing ML Pipeline Best Practices to Scale with LakshmiThejaswi Narasannagari
Mar 16, 2025
LakshmiThejaswi Narasannagari, a seasoned machine learning engineer with experience at Intuit and Poshmark, shares her journey from software development to the ML landscape. She discusses best practices for testing ML pipelines and the importance of automation in ensuring model accuracy. The conversation highlights understanding the distinction between generative AI and predictive analytics, the critical components of a machine learning pipeline, and offers insights on mentorship and resources for those venturing into AI and automation.
AI Snips
Chapters
Transcript
Episode notes
Stay Current and Monitor Models
- Regularly read research papers to understand machine learning models and their behavior.
- Use model observability to monitor performance and potential bias.
What is a Model?
- A model acts like an entity that provides responses based on inputs, aiming to mimic human-like answers.
- Generative AI models provide responses closer to human accuracy than traditional predictive models.
Know Your Model Types
- Learn the basics of different machine learning models even if you're focused on operations.
- Understand how models handle data and their types to build effective pipelines.