From Prompt to Production: Scaling Agentic AI with NVIDIA and Azure
Microsoft Events presents a session on scaling agentic AI in enterprises, featuring Matheen Raza. Attendees will learn how NVIDIA’s AI tools on Azure empower organizations to automate workflows and deploy advanced AI agents.
From Prompt to Production: Scaling Agentic AI with NVIDIA and Azure
Session from Microsoft Ignite 2025
Speakers: Matheen Raza, Microsoft + NVIDIA representatives
Session Level: Intermediate (200)
Session Code: BRKSP477
Session Overview
Agentic AI is reshaping traditional enterprise operations by automating complex business workflows and extracting new value from organizational data. This Microsoft Ignite session examines how NVIDIA and Azure collaborate to provide scalable, practical solutions for agent-powered applications across industries.
Key Topics Covered
- NVIDIA’s Role in Software and Accelerated Computing:
- Explains NVIDIA’s foundational technology for building and running large-scale AI agents
- Integration with Microsoft and Industry Partners:
- Highlights seamless integration of NVIDIA’s AI platforms—Nemotron, NIM, and NeMo Agent Toolkit—on Azure AI
- NVIDIA’s Internal Use Cases:
- Real-world applications of agentic AI for complex tasks like chip design via NVIDIA’s Chip Nemo
- Global Workforce AI Agents:
- How NVIDIA IQ and Agentic AI Blueprint are expanding the presence of AI agents in enterprise environments
- Technical Deep Dives:
- Discussion on agent processing, support for long-context reasoning, and interaction patterns
- Model benchmark results and performance highlights
- Trust and Safety in AI:
- Methods for ensuring content safety and transparency of training data
- Insights into guardrails and model evaluation
- Practical Demonstration:
- Video summarization demo showing analytics and safety monitoring in real-time deployments
What You Will Learn
- The capabilities of NVIDIA’s Nemotron models and NeMo Agent Toolkit on Azure AI
- Enterprise best practices for developing, customizing, and deploying agentic AI
- How to evaluate model performance and ensure safety in production environments
- Integration strategies for agentic AI within business-critical workflows
Chapters (Timeline)
- 0:00 – Introduction: NVIDIA’s role in modern enterprise AI
- 4:29 – Integrating Agentic AI with Microsoft and partners
- 6:35 – Creating internal solutions: NVIDIA Chip Nemo
- 9:43 – Entering the global business workforce: NVIDIA IQ & Blueprint
- 12:22 – Technical discussion: agent processing and context
- 16:42 – Ensuring content safety and data transparency
- 17:00 – Model benchmark comparisons and performance
- 24:37 – Customizing models and setting guardrails
- 29:12 – Live demo: video summarization, analytics, and safety