AI TechOps: Data and Infra for AI Foundation to Scale and Manage 1000s of Cohesive Agents
summary
Claims a concrete blueprint for scaling to thousands of cohesive agents with orchestration layers that let third‑party agents interoperate (example: a host agent mediates Salesforce and Langraph integrations). Describes tooling and patterns: GKE‑based provisioning, vector databases and knowledge graphs for grounding, monitoring and guardrails for security, and hierarchical data access controls. Focuses on data pipelines, model-agnostic inference layers, orchestration topologies and where complexity or feedback loops create surprising operational costs.
AI-written summary · about
chapters (10)
- 0:00AI TechOps intro
- 1:43From programming to Agentic AI
- 3:28Scaling cohesive AI agents
- 5:17Single vs multi-agent
- 7:04Data for intelligent agents
- 9:11GenAI beyond the model
- 11:37Orchestrating agents at scale
- 13:25Customer data journey
- 14:53Debugging agentic systems
- 15:57Security privacy in AI TechOps
related talks
27mAI-Powered ETL Pipeline Orchestration: Multi-Agent Systems in the Era of Generative AI
22mThe Agent Control Plane: Managing Thousands of AI Agents
31mAI Agent Orchestration for Enterprise Arch with AWS Bedrock, Flyte & LangGraph
22mHOW TO BUILD YOUR OWN AI AGENT: THE COMPLETE ARCHITECTURE FOR ENGINEERING TEAMS
18mFrom Reactive to Resilient: Agentic AI
30mFuture of AgenticAI: Architecting the Global Control Plane (Keynote)
31mBuilding, Deploying and Managing AI Agents at Scale
24mAI Agents in Production: Why the Platform Matters