Monte Carlo Data + AI Observability Platform

What do you like best about the product?

Monte Carlo monitors our Redshift warehouse with almost no setup. The automatic table monitors catch freshness, volume, and schema changes across our Kafka and ODS tables, so the Data Engineering team can hear about a broken pipeline before the analysts do. Alerts route to Slack and PagerDuty, which fits cleanly into our on-call rotation. Other teams (Search, Ads, Tracking) can also build their own metric monitors and custom SQL monitors on the same platform, so data quality checks aren’t all on our team. I also like the MCP server: from Claude Code, I can triage an alert, check lineage, and review the monitors without leaving the terminal.

What do you dislike about the product?

Some monitors can get noisy, and tuning them still takes a lot of manual effort. For example, one custom SQL monitor on our tracking events triggered 36 incidents in 30 days, and the team has to keep adjusting thresholds by hand to reduce alert fatigue. Metric monitors on our larger Redshift tables also sometimes time out or fail with errors, and it isn’t always obvious from the UI what caused the failure. It’s also easy to end up with monitors that effectively check nothing. A few of ours still point to tables that no longer exist and show “0 tables,” but there’s no warning to the owner. On top of that, many monitors don’t have a notification channel configured, so an anomaly can sit unnoticed. Better default settings for routing and some kind of cleanup or guardrails would help a lot.

What problems is the product solving and how is that benefiting you?

Monte Carlo has solved our problem of discovering bad data too late. In the past, a stale Kafka load or a schema change in a source table could make it all the way to dashboards and models before anyone on the Data Engineering team noticed. Now we have freshness, volume, and schema monitors on our Redshift tables that alert us in Slack and PagerDuty, which lets us fix the pipeline before analysts and product teams rely on the data. It has also helped spread data quality ownership across the company. The Search, Ads, and Tracking teams monitor their own business metrics—such as ranker scores, ad loss rate, and iOS tracking gaps—so they can catch issues in their own domain without coming to us for custom checks. The end result is less time spent on incident investigation and more trust in the numbers people use to make decisions.

What do you like best about the product?

the integration with Slack is very useful in monitoring our services, likewise the addition of AI agent is very helpful for problem solving and troubleshooting issues.

What do you dislike about the product?

the threshold setting options are somehow limited and do not always allow us to capture our use cases

What problems is the product solving and how is that benefiting you?

it helps us to monitor the quality of our data in Snowflake

What do you like best about the product?

Monte Carlo Delivers, user friendly solution for the observability. I tac be used both from UI and MaC, which is big advantage. The UI is sometimes slow. AI agents takes pretty much time when there is many monitors in the domain. Technical support is fast, direct support from Laura is very satisfying. She has deep technical knowledge with quick rising up and stepping into issue resolution.

What do you dislike about the product?

Loading time, UI performance, the timing for AI agents when domain contains many monitors.

What problems is the product solving and how is that benefiting you?

Monitoring the data in the data pipelines.

What do you like best about the product?

Trend analysis and time series Quality checks

What do you dislike about the product?

UI is not that useful. But with coding agents I am able to create any QC monitors easily

What problems is the product solving and how is that benefiting you?

We can find trend outliers so easily

What do you like best about the product?

It is very helpful to control schema changes or even large additions or deletions. Monitors with AI are so intelligent and provide confidence on "if something happens, I'll be notified". If you configure it properly, you can sleep well :)

The UI is super intuitive and easy to set. You can also link it with pager duty to trigger alarms so it is a good quality advisor.

What do you dislike about the product?

The only thing that I can dislike a little bit is how verbose Monte Carlo is if you configure too much things. If you have a big DWH with hundreds of tables, and you configure hundreds of monitors... the alert channel in slack could be a mess. But it depends more on how good you are setting the correct and needed monitors.

What problems is the product solving and how is that benefiting you?

QA. Basically being sure that we don't have massive insertions without knowing, schema changes.. or even X days without data. It is super helpful in this way.

What do you like best about the product?

Monte Carlo solves our business needs and continues to grow and expand its functionality, especially with AI agents. I also like the unified search: I can search in one toolbar and find what I’m looking for, which used to be a pain and still is in some other tools.

What do you dislike about the product?

I still have issues with integrations and lineage tracing, even after including the required tools and getting the right permissions. I can’t see the full data lineage in one graph, including Spark jobs, Airflow DAGs, and datasets. I’d also like lineage to integrate with GitHub without needing other paid third-party tools, and for the agents to query Databricks directly using lineage and alert definitions.

What problems is the product solving and how is that benefiting you?

I use Monte Carlo to monitor bad data.

What do you like best about the product?

The combination of automated anomaly detection and end-to-end data lineage has been fantastic for us. It proactively alerts us when data is missing or looks anomalous, which has significantly reduced our overall data downtime and strengthened the organization’s trust in our analytics.

What do you dislike about the product?

Everything has been great so far. I don't have any negative feedback at this time, but I'll update my review if I run into any issues.

What problems is the product solving and how is that benefiting you?

We used to struggle with manual data quality testing that was slow and incomplete. Now, we use Monte Carlo, which performs continuous, automated monitoring across our stack with zero noticeable performance impact on our warehouse. This has resulted in faster root cause analysis, fewer broken dashboards, and a much happier data engineering team.

What do you like best about the product?

What I like the most about Monte Carlo is its UX as it's very easy to use even for the first time. Its integrations, i.e., with Slack, work really smoothly and perform quickly. I also enjoy its AI features that help me with everyday work with MC as well as the support from the company. The money paid is worth it.

What do you dislike about the product?

Like I said before, Monte Carlo is an amazing tool, but if I were to say what they could improve, it would be to make the alerts searching a bit easier.

What problems is the product solving and how is that benefiting you?

The problem Monte Carlo solves the best is alerting on data quality in our warehouse.

What do you like best about the product?

I love the monitors-as-code feature, since it lets us keep our monitors directly in our codebase. The integrations with the data warehouses and platforms we already use are also stellar.

What do you dislike about the product?

Ironically, the main pain point for me is the monitors-as-code data structure. It was recently revamped, and the new version is definitely clearer, but the overall structure still feels a bit unintuitive, and the documentation can be lacking at times.

Another issue for us is that we can’t define cron schedules for certain monitors. That makes it problematic to monitor our pipelines that themselves run on cron jobs. I understand Monte Carlo prioritizes and pushes the auto-ml / AI monitors, but sometimes simpler monitors that run on exact schedules are simpler and, for our use case, superior—which is what we prefer.

What problems is the product solving and how is that benefiting you?

Completeness, volume, and uniqueness checks across many of our products require a solution that’s easy to use, simple to set up, and straightforward to integrate, with monitoring and alerting built in. Monte Carlo really helps here.

What do you like best about the product?

I like that Monte Carlo monitors changes to my data and helps me take action proactively. I can identify data completeness issues before the business finds out and an incident is created.

What do you dislike about the product?

I get repeat alerts for the same pattern even after I investigate and mark an issue as expected. For example, if I resolve an alert for 1,000 rows deleted, I don’t want another alert when 900 rows are removed. I’d like Monte Carlo to adjust the threshold and start monitoring above 1,000 instead of 100, using learned metrics.

What problems is the product solving and how is that benefiting you?

I use Monte Carlo to spot data completeness issues proactively, before the business finds out and raises an incident.