Practical AI Operationalizing ML/AI with MemSQL
Jun 29, 2020
Nikita Shamgunov, Co-CEO of MemSQL and expert in data management, discusses the challenges of deploying AI models in production. He highlights the integration of machine learning into SQL workflows and how MemSQL enhances performance for real-time applications like fraud detection. Nikita shares insights on data management evolution, the significance of distributed databases, and the vital role of AI in improving operational workloads. He also reflects on how MemSQL's solutions aided healthcare during the COVID-19 pandemic, showcasing data fairness and community collaboration.
AI Snips
Chapters
Transcript
Episode notes
Career Shift
- Nikita Shamgunov transitioned from research to database engineering at Microsoft SQL Server.
- This shift exposed him to user-focused development, performance, and cloud transitions.
Distributed Systems
- Distributed systems are essential for high-performance workloads, which motivated the creation of MemSQL.
- Facebook's reliance on distributed systems validated this approach.
Single Pane of Glass
- Modern workloads require scaling storage and compute for both low-latency operations and large analytical queries.
- MemSQL aims to be a single, serverless SQL interface for all data and workloads.