- Datadog (DDOG, Financial) has unveiled two groundbreaking AI research projects, Toto and BOOM, to enhance observability forecasting.
- Toto is an open-source time series foundation model offering zero-shot forecasting for anomaly detection and capacity planning.
- BOOM is the largest public benchmark for observability metrics with 350 million observations across 2,807 multivariate series.
On May 21, 2025, Datadog, Inc. (DDOG) introduced two significant AI research initiatives through its Datadog AI Research division, with the launch of Toto and BOOM. These projects aim to advance observability forecasting in the research community under a permissive open-source license.
Toto is positioned as the first open-source time series foundation model (TSFM) designed specifically for observability data. It leverages Datadog's internal telemetry metrics to offer zero-shot forecasting capabilities, enabling instant anomaly detection and capacity planning. This advancement marks a significant milestone by outperforming existing TSFMs in terms of observability data analysis.
In addition, BOOM establishes the largest public benchmark of observability metrics, encompassing 350 million observations across 2,807 real-world multivariate series. This benchmark addresses unique challenges such as scale, sparsity, spikes, and cold-start issues intrinsic to production telemetry data, providing a valuable resource for researchers in the field.
According to Chief Scientist Ameet Talwalkar, these initiatives signify the commencement of continuous AI project releases from Datadog AI Research, with a strategic aim of collaborating with applied AI teams to transform theoretical advances into practical solutions for customers.
Both Toto and BOOM are now available for immediate download, inviting the research and open-source communities to utilize and enhance these tools for observability forecasting.
