Stay current on new cloud trends. Top software companies, respected industry analysts, and experienced consultants join Amazon Web Services leaders to talk about the cloud topics that matter to you—including the latest in AI, migration, Software-as-a-Service, and more. We produce new episodes regularly.
Kui Jia, Sumo Logic's Vice President of Engineering and Head of AI, shares how their AWS-powered AI agents transform chaotic security investigations into streamlined workflows.
Topics Include:
- Kui Jia leads AI Engineering at Sumo Logic
- SREs and SOC analysts work under chaotic, high-pressure conditions
- Teams constantly switch between different vendor tools and platforms
- Investigation requires quick hypothesis formation and complex query writing
- Sumo Logic processes petabytes of data daily across enterprises
- Company serves 2,000+ enterprise customers for 15 years
- Platform focuses on observability and cybersecurity use cases
- Investigation journey: discover, diagnose, decide, act, learn phases
- Data flows from ingestion through analytics to human insights
- Traditional workflow relies heavily on tribal domain knowledge
- Senior engineers create queries that juniors struggle to understand
- War room situations demand immediate answers, not learning curves
- Context switching between tools wastes time and creates friction
- Multiple AI generations deployed: ML anomaly detection to GenAI
- Agentic AI enables reasoning, planning, tools, and evaluation capabilities
- Mo Copilot launched at AWS re:Invent as AI agent suite
- Natural language converts high-level questions into Sumo queries
- System provides intelligent autocomplete and multi-turn conversations
- Insight agents summarize logs and security signals automatically
- Knowledge integration combines foundation models with proprietary metadata
- AI generates playbooks and remediation scripts for automated actions
- Three-tier architecture: Infrastructure, AI Tooling, and Application layers
- Built on AWS Bedrock with Nova models for performance
- Focus on reusable infrastructure and AI tooling components
- Data differentiation more important than AI model selection
- Golden datasets and contextualized metadata are development challenges
- Guardrails and evaluation frameworks critical for enterprise deployment
- AI observability enables debugging and performance monitoring
- Enterprise agents achievable within one year development timeline
- Future vision: multiple AI agents collaborating with human investigators
Participants:
Further Links:
See how Amazon Web Services gives you the freedom to migrate, innovate, and scale your software company at https://aws.amazon.com/isv/
