AI People Wiki

Builders & open source

Abdul Fatir Ansari

Senior Applied Scientist, AWS AI, Amazon

Abdul Fatir Ansari is a Senior Applied Scientist at Amazon Web Services, working on time series forecasting, log analytics and generative models, and in particular on general-purpose pretrained models for time series. Ansari is the first author of the Chronos (2024) and Chronos-2 (2025) papers, which introduced Amazon's open pretrained time series forecasting models, and contributes to AWS's open-source AutoGluon and GluonTS libraries.123

At a glance

Current role
Senior Applied Scientist, AWS AI
Company
Amazon
Sector
Company
Work listed
2
Sources cited
5
Profile reviewed

Career

Ansari's site lists AWS AI in 2021 and from 2022 to the present, and describes Ansari as a Senior Applied Scientist at AWS working on time series forecasting, log analytics and generative models, with recent work on general-purpose pretrained models for time series.1

Research

Ansari's research covers time series analysis and generative models, including probabilistic generative modeling, unsupervised learning and representation learning. As first author of Chronos (2024), Ansari described a framework that scales and quantizes time series into a fixed vocabulary and trains T5-based models on them, and as first author of Chronos-2 (2025) extended the approach to multivariate and covariate-informed forecasting.123

Ansari also co-authored fev-bench (2025), a benchmark for time series forecasting with Oleksandr Shchur and colleagues at AWS, and in August 2025 lectured on building foundation models for time series at the Oxford ML School.41

Work and contributions

  • Chronos Open weights

    First author of "Chronos: Learning the Language of Time Series" (2024), which tokenizes time series values and trains transformer language model architectures on them to build pretrained probabilistic forecasting models.25

  • Chronos-2 Open weights

    First author of "Chronos-2: From Univariate to Universal Forecasting" (2025), describing the Chronos-2 model with zero-shot support for univariate, multivariate and covariate-informed forecasting.351

Timeline

  1. Began the AWS AI role that Ansari's site lists as running from 2022 to the present.1
  2. Posted "Chronos - Learning the Language of Time Series" as first author.2
  3. Delivered a lecture on building foundation models for time series at the Oxford ML School.1
  4. Co-authored fev-bench, a benchmark for time series forecasting.4
  5. Posted the Chronos-2 paper as first author, alongside the model's release.35

Where they work

Suggest a correction or removal

Profiles cover public professional work only, and each statement cites a source. If something is out of date, missing or wrong, tell us and link a source. People can also ask for their profile to be corrected or removed.

References

5 sources · profile reviewed Oct 22, 2026 · data updated Oct 22, 2026

  1. Abdul Fatir Ansari — homepageAbdul Fatir Ansari · Official page · accessed Oct 22, 2026
  2. Chronos: Learning the Language of Time Series (arXiv 2403.07815)arXiv · Paper · published Mar 12, 2024 · accessed Oct 22, 2026
  3. Chronos-2: From Univariate to Universal Forecasting (arXiv 2510.15821)arXiv · Paper · published Oct 17, 2025 · accessed Oct 22, 2026
  4. fev-bench: A Realistic Benchmark for Time Series Forecasting (arXiv 2509.26468)arXiv · Paper · published Sep 30, 2025 · accessed Oct 22, 2026
  5. amazon-science/chronos-forecasting READMEamazon-science/chronos-forecasting (GitHub repository) · Repository · accessed Oct 22, 2026