profile picture

thinking about distributed, efficient and collaborative ml 🤔

ML Research @ University of Cambridge & Pluralis Research.

Updates

  • 📅 September 2026: Started as a research scientist intern at Pluralis Research.
  • 🏅 July 2026: Received ICML 2026 Gold Reviewer Award.
  • 📰 🇰🇷 July 2026: LorDO accepted to ICML 2026. See you in Seoul.
  • 📰 🇧🇷 April 2026: MT-DAO and DES-LOC accepted to ICLR 2026. See you in Rio.
  • 📰 🇩🇪 March 2026: Panza accepted to CPAL 2026. See you in Tubingen.

Research

My research focuses distributed machine learning, foundation-model training, machine learning efficiency, and efficient optimization. I am particularly interested in designing methods that make large-scale training more communication- and compute-efficient, and more robust in realistic deployment settings.

Distributed training Efficient and scalable learning across geo-distributed devices and clusters.
Machine learning efficiency Reducing communication, compute, and memory costs while preserving model quality.
Foundation models Training and adaptation of large models under practical constraints.
Optimization Algorithms that improve convergence, efficiency, and robustness.

please see my Google Scholar profile for an up-to-date list of my publications.

education

experience

  • Pluralis Research logo

    Sep 2026 - Present

    research scientist intern | pluralis research

    Working on the intersection of distributed and decentralized training and architecture design for foundation models.

  • CaMLSys logo

    Oct 2025 - Present

    research assistant | university of cambridge

    Developing novel local-update optimization algorithms for large-scale federated foundation model training to improve convergence rates and communication efficiency while maintaining performance.

  • flower labs logo

    May 2025 - May 2026

    research scientist intern | flower labs

    Worked on Flower's model training team across pre-training and post-training of LLMs, developing distributed training methods with a focus on optimizers and improvements to MoE architectures.

  • DAS logo

    Oct 2024 - March 2025

    research assistant | institute of science and technology austria Alistarh Group

    Working as a research assistant under the supervision of Dan Alistarh in the Distributed Algorithms and Systems (DAS) group focusing on compressed federated (pre)training of foundation models.

  • Summer 2023 - Summer 2023

    app. ai research intern | jp morgan chase & co. (Applied Innovation of Artificial Intelligence (AI2))

    Developed unsupervised learning methods, with an associated parallel data processing pipeline, for anomaly detection which increased previous performance by 30%.

  • Summer 2022 - Summer 2022

    ml eng intern | jp morgan chase & co. (Applied Innovation of Artificial Intelligence (AI2))

    Pursued personal research in Natural Language Processing, building a search engine, powered by dense passage retrieval, with q&a capabilities. Assisted researchers in delivering proof of concept projects, onboarding them to CI/CD platforms.

academic service

10/2025 - present supervision university of cambridge

  • shrey biswas part ii thesis on model merging
  • francesco simioni masters thesis on latent representations in federated learning (accepted at MobiUK)

2/2025 - present reviewing

  • journals tmlr
  • conferences neurips 2026, icml 2026 (Gold Reviewer Award)
  • workshops neurips 2026 axiom, neurips 2026 codec-fm, icml 2026 adaptfm (technical program chair), aaai 2026 flca, iclr 2025 mcdc

volunteering

02/2023 - present ml systems & theory co-lead

  • At the Cohere Labs open-science community, I have the privilege of being one of the co-leads for the ML Systems & Theory group. This is designed to provide a platform for exploring research ideas and open questions at this intersection.

Collaboration

If you are interested in collaborating or have questions about my research, please feel free to get in touch.