Monte Carlo integrates with Agent Bricks for Databricks observability
Monte Carlo has launched an integration with Agent Bricks on Databricks to extend observability across the full data stack, covering Delta Lake, Lakeflow, and AI agents. This unified view enables enterprises to trace failures and validate data reliability from raw data to agent actions. The solution is immediately available for Databricks customers.
*this image is generated using AI for illustrative purposes only.
Monte Carlo today announced its integration with Agent Bricks, Databricks' platform to build, deploy, and govern AI agents on enterprise data. This development extends Monte Carlo’s observability capabilities to the agent layer, providing enterprises with a continuous, unified view across the full Databricks Data Intelligence Platform. The integration aims to help organizations distinguish between data, model, and pipeline failures to ensure the reliability of AI agents in production.
Enterprises utilizing Databricks rely on monitoring to maintain the health of data underlying analytics and AI. The integration connects three interconnected layers of the stack. The first layer, Delta Lake & Data Tables, offers continuous monitoring for freshness, schema drift, volume anomalies, and quality degradation. The second layer, Lakeflow, provides health monitoring, anomaly detection, and end-to-end lineage across data engineering workflows. The third layer, Agent Bricks, delivers observability across tool calls, retrieval steps, model interactions, orchestration workflows, and data inputs.
Unified Observability Layers
The integration creates a comprehensive audit trail from raw data in Delta Lake to actions taken by deployed agents. This structure allows engineering teams to trace failures, validate data reliability, and identify root causes of agent issues.
| Layer | Function |
|---|---|
| Delta Lake & Data Tables | Monitors freshness, schema drift, volume anomalies, and quality degradation. |
| Lakeflow | Tracks health, anomaly detection, and end-to-end lineage in data engineering. |
| Agent Bricks | Provides observability for tool calls, retrieval steps, and model interactions. |
Barr Moses, co-founder and CEO of Monte Carlo, emphasized the necessity of visibility in the new infrastructure layer. "Deploying agents in production means managing an entirely new layer of infrastructure — and most enterprises have no visibility into it," said Moses. "Databricks customers now have a single, cohesive view of everything their agents run on and everything their agents do."
Michael Weiss, AVP of Product Management at Nasdaq, highlighted the importance of data trust. "Even if you have access to all of the information in your entire data ecosystem, if you can't trust the data, then it's no good," said Weiss. The integration is available now for enterprises running on the Databricks Data Intelligence Platform.
Historical Stock Returns for Monte Carlo Fashions
| 1 Day | 5 Days | 1 Month | 6 Months | 1 Year | 5 Years |
|---|---|---|---|---|---|
| -1.05% | +2.09% | +0.68% | -20.67% | -7.50% | +64.17% |
How will this unified observability impact the speed at which enterprises can identify and resolve AI agent failures in production?
Will this integration drive increased adoption of Databricks' Agent Bricks among enterprises hesitant to deploy AI agents due to reliability concerns?
Could this partnership pressure other data platform providers to develop similar end-to-end observability solutions for their AI agent ecosystems?
20
Access Market Heat-map & Insights 
Create, Save & Share your Screener 
Use 200+ Screening Filters 
