Matthew Stewart

Matthew Stewart

Lead AI Researcher

Pelago Health

mpstewart94@gmail.com

Biography

I am Lead AI Researcher at Pelago Health, where I work on AI for behavioural healthcare. My research focuses on how intelligent systems can be evaluated, governed, and held accountable in the settings where people use them. I previously worked as a Postdoctoral Researcher in Harvard University’s Edge Computing Lab, following my PhD at Harvard.

My work connects responsible AI and algorithmic accountability with machine learning systems, intelligent sensing, and digital health. My sole-authored paper, Beyond Explanation: Evidentiary Rights for Algorithmic Accountability, received a Best Paper Award at ACM FAccT 2026. My collaborative research includes machine learning sensors, TinyML, and benchmarks for robotics and neuromorphic computing.

I also care about making machine learning education accessible. I co-created the HarvardX TinyML Professional Certificate and contribute to the open-source Machine Learning Systems textbook. Alongside my research, I serve on editorial boards and contribute to research assessment and professional communities. Outside work, I enjoy endurance sport, including triathlons and long-distance running.

Interests

  • Responsible AI & algorithmic accountability
  • AI evaluation & governance
  • Digital health & human–AI interaction
  • Machine learning sensors & TinyML
  • Benchmarking intelligent systems

Education

  • PhD in Engineering Sciences and Data Science, 2023

    Harvard University

  • MS in Engineering Sciences, 2020

    Harvard University

  • Integrated BEng/MEng in Mechanical Engineering, 2017

    Imperial College London

Recent News

[June 2026] My paper Beyond Explanation received a Best Paper Award at ACM FAccT 2026.

[2026] New journal publications include Datasheets for Machine Learning Sensors in AI Magazine and Wake Vision in the Journal of Data-centric Machine Learning Research.

[September 2025] I joined Pelago Health as Lead AI Researcher.

Work Experience

Pelago Health

Lead AI Researcher · September 2025–present

Research and development of AI for behavioural healthcare, with an emphasis on evaluation, safety, and clinician oversight.

Harvard University — Edge Computing Lab

Postdoctoral Researcher · January 2023–April 2024

Research on machine learning sensors, embedded intelligence, and benchmarking for robotics and neuromorphic computing.

Featured Publications

A selection of my research. See Google Scholar for the full publication list.

More selected publications and preprints
  • Wake Vision: A Tailored Dataset and Benchmark Suite for TinyML Computer Vision Applications. C. Banbury et al., including M. Stewart. Journal of Data-centric Machine Learning Research, 2026. Paper.
  • The NeuroBench framework for benchmarking neuromorphic computing algorithms and systems. J. Yik et al., including M. Stewart. Nature Communications, 2025. Article.
  • A2Perf: Real-World Autonomous Agents Benchmark. I. Uchendu et al., including M. Stewart. arXiv preprint, 2025. Preprint.
  • RobotPerf: An open-source, vendor-agnostic, benchmarking suite for evaluating robotics computing system performance. V. Mayoral-Vilches et al., including M. Stewart. ICRA, 2024. Paper.
  • Machine Learning Sensors: A Design Paradigm for the Future of Intelligent Sensors. P. Warden, M. Stewart, B. Plancher, S. Katti, and V. Janapa Reddi. Communications of the ACM, 2023. Article.
  • Is TinyML Sustainable? S. Prakash, M. Stewart, C. Banbury, M. Mazumder, P. Warden, B. Plancher, and V. Janapa Reddi. Communications of the ACM, 2023. Article.
Accepted work on agentic AI
  • Making AI Memory Transparent: Explainability for Persistent State in Agentic Systems. M. Stewart. EXTRAA-MAS @ AAMAS 2026, Springer LNAI. Accepted for publication.
  • Ethical Non-Closure Under Agentication: Why Action-Level Frameworks Fail for Trajectory-Producing Systems. M. Stewart. COINE 2026, Springer LNAI. Accepted for publication. Workshop paper.

Education

PhD in Engineering Sciences and Data Science

Harvard University

August 2017–January 2023

Kennedy Research Scholar.

MS in Engineering Sciences

Harvard University

August 2017–May 2020

Integrated BEng/MEng in Mechanical Engineering

Imperial College London

October 2013–June 2017

First Class Honours; year abroad at the National University of Singapore.

Teaching

HarvardX TinyML Professional Certificate

Co-creator and Course Developer · 2020–2022

A four-course sequence introducing applied machine learning on embedded devices. Read about the education programme.

CS249r: Tiny Machine Learning

Head Teaching Assistant · Harvard University, 2023

ES112: Engineering Thermodynamics by Case Study

Head Teaching Assistant · Harvard University, 2018–2021

I also contribute to the Machine Learning Systems textbook and mentor through MLCommons Rising Stars and Harvard’s alumni community.

For accessible introductions to machine learning and its implications, see my public writing.

Selected Awards

  • ACM FAccT Best Paper Award (2026), for Beyond Explanation: Evidentiary Rights for Algorithmic Accountability.
  • EDSAFE Catalyst Fellow, EDSAFE AI Alliance (2026).
  • Kennedy Research Scholarship, Kennedy Memorial Trust (2021).
  • AAAR Student Poster Award, American Association for Aerosol Research (2019).
  • Certificates of Distinction in Teaching, Harvard Derek Bok Center (2018, 2019, 2021).

Professional Service & Affiliations

Editorial service

  • Associate Editor, IEEE Transactions on Neural Networks and Learning Systems (2026–present).
  • Associate Editor, IEEE Internet of Things Journal (2026–present).

Research assessment & professional service

  • Innovate UK assessor; EPSRC Peer Review College member; Horizon Europe expert evaluator.
  • CSAB/ABET Data Science Program Evaluator; Science Council assessor.
  • Program committees for AIES 2026, IJCAI–ECAI 2026, and IEEE SysCon 2026.
  • INFORMS Committee on Industry–Academia Collaborations; Royal Statistical Society Data Ethics & Governance Section; IEEE Humanitarian Technologies Board Tech4Good Committee.

Professional recognition

Fellow of the British Computer Society (FBCS), the Institute of Mathematics and its Applications (FIMA), the Institute of Analytics (FIoA), and the Institution of Engineers in Scotland (FIES).

Chartered Mathematician (CMath), Chartered Scientist (CSci), Chartered Statistician (CStat), Chartered IT Professional (CITP), and ASA Accredited Professional Statistician (PStat).