Yasitha Liyanage

Yasitha Liyanage

Senior Data and Applied Scientist

Microsoft Corporation

Biography

Yasitha Liyanage is a Senior Data and Applied Scientist at Microsoft, working on machine learning, cloud observability, and causal inference. He builds large-scale anomaly detection and automated outage attribution systems for Azure cloud infrastructure.

He received his PhD in Electrical and Computer Engineering from the University at Albany, State University of New York, in January 2022. His doctoral research focused on dynamic instance-wise decision-making for machine learning. Previously, he worked on credit risk and revenue forecasting as an Applied Scientist Intern at Amazon.

Research Interests

  • Machine Learning & AI
  • Cloud Observability
  • Signal Processing
  • Causal Inference

Education

  • PhD in Electrical & Computer Engineering, January 2022

    University at Albany, State University of New York (SUNY)

  • BSc in Electrical & Electronic Engineering, October 2016

    University of Peradeniya, Sri Lanka

Experience

 
 
 
 
 

Senior Data and Applied Scientist

Microsoft Corporation

March 2024 – Present Redmond, Washington
 
 
 
 
 

Data and Applied Scientist–II

Microsoft Corporation

Jan 2022 – March 2024 Redmond, Washington
 
 
 
 
 

Applied Scientist Intern

Amazon Capital Services, Inc

May 2021 – Aug 2021 Seattle, WA
 
 
 
 
 

Applied Scientist Intern

Amazon Capital Services, Inc

Jun 2020 – Aug 2020 Seattle, WA
 
 
 
 
 

Graduate Research Assistant

The State University of New York at Albany

Aug 2017 – Dec 2021 Albany, New York
 
 
 
 
 

Graduate Research Assistant

University of Peradeniya

Nov 2016 – Jul 2017 Peradeniya, Sri Lanka

Recent Publications

Quickly discover relevant content by filtering publications.

Causal Attribution of Azure Control-Plane Anomalies

MLADS 2026 · Redmond, WA · June 8–11, 2026

Aegis: Attribution of Control Plane Change Impact across Layers and Components for Cloud Systems

Dynamic Feature Selection for Classification in Structured Environments

Dynamic Instance-Wise Classification in Correlated Feature Spaces

Dynamic Instance-Wise Joint Feature Selection and Classification

Near Real-Time Freeway Accident Detection

In this paper, the problem of detecting accidents using speed sensors distributed spatially on a freeway is considered. Due to the …

Sequential Heterogeneous Feature Selection for Multi–Class Classification: Application in Government 2.0

Herein, the problem of multi-class classification of participatory civil issue reports in crowdsourcing platforms is addressed. …

Patents

Real-Time Detection of Latency Anomalies Affecting Cloud Production Workloads

Y. Liyanage, K. Hsieh, M. Chintalapati, V. Ramesh, S. Zuehlsdorff

Awards

Microsoft Patent Award

Azure Core Champ Award

Distinguished Doctoral Dissertation Award

Graduate Student Association Grant Award

NSF Student Travel Grant to attend IEEE ICASSP 2020

President’s Award for Scientific Research 2019, Sri Lanka

IEEE Signal Processing Society Travel Grant to attend GlobalSIP 2018

Full Colors, Colors Awarding Ceremony 2016, University of Peradeniya, Sri Lanka

Fun Facts About Me

Outside of work, I enjoy volleyball, cycling, hiking, and reading.

Volleyball

Cycling

Hiking

Reading

Contact

  • yasithashehanliyanage (at) gmail (dot) com
  • Chicago, Illinois, USA