The integration of Venn into the company’s Flow, Observe, and Fleet products enables engineers to author pipeline rules and manage fleet agents conversationally. To address concerns regarding automated system changes, Apica implemented a mandatory human-in-the-loop requirement for any action that could materially alter configuration or delete data streams. Administrators must explicitly approve destructive operations, such as disabling collectors, before they execute.
Apica Upgrades Ascent Platform to Manage Ballooning AI Telemetry
As AI agents trigger a 9.5x surge in telemetry volumes, Stockholm-based Apica is shifting infrastructure management from manual scripting to natural language. The company’s Ascent 3.0 update introduces a persistent AI assistant, Venn, and an MCP server to allow enterprise teams to automate pipeline configurations while maintaining human oversight.

For organizations already invested in proprietary AI tools, the new built-in Model Context Protocol (MCP) server facilitates a "bring your own AI" (BYOAI) approach. This allows external assistants to inspect and manage pipelines and alerts directly without the need to route sensitive telemetry through third-party vendors. Complementing these control features is a new flow-only mode that bypasses Lake storage overhead. During internal engineering benchmarks, this configuration moved telemetry data up to 10x faster, providing necessary bandwidth for the heavy data loads generated by autonomous agents. Andi Mann, CPTO at Apica, emphasized that these updates prioritize data sovereignty, ensuring customers retain ownership of their telemetry and the freedom to select their own AI providers.


Comments (0)
No comments yet. Be the first!