I’ve been exploring how Model Context Protocol (MCP) could change the way engineers interact with observability data, particularly logs stored in Dynatrace Grail.
Instead of always starting with dashboards or manually constructing queries, imagine an AI assistant being able to use an MCP-based integration to retrieve the relevant Grail data, work with DQL, and help engineers investigate operational issues through natural-language interactions.
Some areas I’m particularly interested in:
- Using MCP to securely expose relevant Grail observability data to AI assistants
- Natural language to DQL workflows
- AI-assisted log investigation and troubleshooting
- Correlating logs with application and infrastructure context
- Providing appropriate access controls and guardrails for AI-driven queries
For me, the exciting part isn’t simply letting AI “search logs.” It’s about combining Grail’s observability context with AI-driven investigation while maintaining enterprise security and governance.
I’d be interested to hear from the community: Is anyone experimenting with MCP around Dynatrace Grail or DQL today? What use cases and challenges are you seeing?
#Dynatrace grail #MCP observability dql #GenAI aiops