---
title: "Datadog at NeurIPS 2025"
description: "Datadog AI Research is attending NeurIPS 2025 in San Diego as a Gold sponsor, presenting work on time series foundation models and observability-native AI."
author: "Ameet Talwalkar, Emaad Khwaja, Ben Cohen"
date: 2025-12-01
tags: ["bits ai", "watchdog", "ai", "observability", "ai research"]
blog_type_id: ai
locale: en
---

Datadog AI Research is excited to participate in [NeurIPS 2025](https://neurips.cc/) as Gold-level sponsor. As one of the most important conferences for advancing machine learning research, NeurIPS is a natural fit for our work on observability-native foundation models and AI agents.

If you're attending in San Diego, we'd love to talk about time series foundation models, incident automation, and intelligent code repair. Here's where you can find us:

## Poster session

[**This Time is Different: An Observability Perspective on Time Series Foundation Models**](https://www.datadoghq.com/blog/ai/toto-boom-unleashed.md)

-**San Diego poster session**

-**Location**: Exhibit Hall C, D, E — Poster #2310

-**Date and time**: Friday, December 5, 16:30–19:30 PST

In this work, we introduce Toto, a time series foundation model tailored to observability metrics, and BOOM, a large-scale benchmark built from real-world observability workloads.

## Visit us in the Expo Hall

We'll be at **Booth 631** in the San Diego Convention Center Expo Hall from **Tuesday, December 2 to Friday, December 5, 2025**. Drop by to learn how we're building observability-native AI—from time series foundation models to autonomous SRE agents and AI-powered developer tools.

## Workshop: Recent Advances in Time Series Foundation Models (BERT²S)

We're also participating in the [Recent Advances in Time Series Foundation Models: Have We Reached the “BERT Moment”? (BERT²S)](https://berts-workshop.github.io/) workshop, held on **Sunday, December 7, 2025**, from **08:00–17:00 PST**. The workshop focuses on benchmarking, scaling laws, and the real-world impact of time series foundation models.

Ameet Talwalkar, Associate Professor at CMU and Chief Scientist at Datadog, and Chenghao Liu, Staff Research Scientist, will join the panelists to discuss the role of observability-native datasets, strong supervised baselines, and practical evaluation in getting from “paper SOTA” to robust production systems.

## Work with us

Datadog AI Research focuses on high-risk, high-reward projects rooted in cloud observability and security, building on a history of AI-powered features like [**Watchdog**](https://www.datadoghq.com/product/platform/watchdog/), [**Bits AI**](https://www.datadoghq.com/product/platform/bits-ai/), and [**Toto**](https://github.com/DataDog/toto).

If you'd like to work with us on foundation models like Toto, build benchmarks like BOOM, and help define the next generation of AI agents for operations and code repair, come find us at NeurIPS and explore the AI Research Scientist and AI Research Engineer roles on the [**Datadog Careers site**](https://careers.datadoghq.com/all-jobs/?s=%22Datadog%20AI%20Research%22&utm_source=engblog&utm_medium=corpsite&utm_campaign=engcommunity-2025-ai-bitsai&gh_src=hm4uekgj1us).