Evolving MLOps Platforms for Generative AI and Agents with Abhijit Bose - #714
Published:Jan 13, 2025 22:25
•1 min read
•Practical AI
Analysis
This podcast episode from Practical AI features Abhijit Bose, head of enterprise AI and ML platforms at Capital One, discussing the evolution of their MLOps and data platforms to support generative AI and AI agents. The discussion covers Capital One's platform-centric approach, leveraging cloud infrastructure (AWS), open-source and proprietary tools, and techniques like fine-tuning and quantization. The episode also touches on observability for GenAI applications and the future of agentic workflows, including the application of OpenAI's reasoning and the changing skillsets needed in the GenAI landscape. The focus is on practical implementation and future trends.
Key Takeaways
- •Capital One is evolving its MLOps and data platforms to support generative AI and AI agents.
- •The company leverages cloud infrastructure (AWS) and a mix of open-source and proprietary tools.
- •The discussion covers practical aspects like fine-tuning, quantization, and observability for GenAI applications.
Reference
“We explore their use of cloud-based infrastructure—in this case on AWS—to provide a foundation upon which they then layer open-source and proprietary services and tools.”