Postdoctoral Associate (Resource-Efficient Machine Learning)

Singapore-MIT Alliance for Research and Technology Centre; Singapore; On-site; SGD 60,000 - 80,000; Full time

Job summary

A leading research institute in Singapore is seeking a Postdoctoral Associate to advance resource-efficient machine learning. This role offers a 1-year appointment focusing on developing next-generation methods to enhance computational efficiency in large-scale AI systems. Candidates should have a doctoral degree in a quantitative field, experience with deep learning models, and a track record of research publication. The position requires collaboration with faculty from MIT and NUS, fostering an interdisciplinary research environment.

Qualifications

• Proven track record in high-quality research contributions published in top-tier conferences or journals.
• Experience in training, fine-tuning, or deploying large-scale deep learning models.
• Ability to collaborate effectively within multidisciplinary teams.

Responsibilities

• Contribute to resource-efficient machine learning methods improving computational efficiency.
• Design and implement next-generation algorithms and architectures.
• Collaborate with research staff and publish results in top-tier venues.

Skills

Large-scale deep learning models

Modern ML frameworks (e.g., PyTorch, JAX)

High-performance computing

Independent research skills

Education

Doctoral degree in Computer Science, Machine Learning, or related discipline

Tools

Transformers

Job description

Project Overview

We are hiring highly motivated and talented Postdoctoral Associates who are interested in advancing the state of the art in resource-efficient machine learning at the Singapore-MIT Alliance for Research & Technology (SMART). This position is part of the program: “Mens, Manus, and Machina: How AI empowers people and the city in Singapore (M3S).” The role offers a 1-year appointment with the possibility of renewal/extension, focusing on establishing a new paradigm for resource-efficient machine learning that balances computational efficiency with state-of-the-art performance.

Research Focus: Foundation Models & Next-Generation Methods

As the field increasingly pivots toward foundation models, efficiency has become a central challenge. Addressing this challenge requires approaches that go beyond incremental optimization. We seek researchers to develop next-generation machine learning methods that fundamentally rethink how large-scale AI systems are trained, fine-tuned, and deployed. Our focus is on the development and application of advanced techniques, including AutoML, Bayesian optimization, neural architecture search, reinforcement learning, and active learning, with the explicit goal of achieving significant gains in computational efficiency without sacrificing performance.

Key Responsibilities

• Contribute to the development of resource-efficient machine learning methods that improve computational efficiency of large-scale AI systems, such as foundation models (e.g., LLMs, VLMs, Multimodal models) and World models, while maintaining or improving performance.
• Design and implement next-generation algorithms, architectures, and learning strategies that fundamentally challenge existing resource constraints in large-scale AI systems.
• Prototype, implement, and rigorously evaluate complex machine learning systems to assess the performance, scalability, and practical feasibility of novel theoretical ideas.
• Collaborate with other research staff and students to publish research results in top-tier conferences and journals, focusing on venues associated with the above-mentioned areas.
• Work closely with and provide technical and intellectual mentorship to PhD students and research engineers within the SMART and M3S research ecosystem, fostering a collaborative and interdisciplinary research environment.

Requirements

• A doctoral degree in Computer Science, Machine Learning, Statistics, Mathematics, or a closely related quantitative discipline.
• Demonstrated experience with large-scale deep learning models and modern ML frameworks (e.g., PyTorch, JAX, Transformers), including training, fine-tuning, or deployment at scale.
• A proven track record of high-quality research contributions published in top-tier machine learning conferences or journals.
• Proficiency in high-performance computing, distributed and parallel training, and end-to-end modern machine learning pipelines.
• Ability to conduct independent research, critically engage with emerging challenges in AI efficiency and sustainability, and collaborate effectively within multidisciplinary teams.

The postdoctoral fellow will be based at the SMART Centre in Singapore. You will be working directly with the following leading faculty from MIT and NUS:

Daniela Rus (MIT) https://danielarus.csail.mit.edu/

Armando Solar-Lezama (MIT) https://people.csail.mit.edu/asolar/

Bryan Kian Hsiang Low (NUS) https://www.comp.nus.edu.sg/~lowkh/index.html

To apply, please visit our website at: https://portal.smart.mit.edu/careers/career-opportunities

Interested applicants are invited to send in their full CV/resume, cover letter and list of three references (to include reference names and contact information). We regret that only shortlisted candidates will be notified.