Machine Learning Engineer

About Insilico Medicine

Insilico Medicine is an end-to-end, artificial intelligence (AI) -driven pharma-biotechnology company with a mission to accelerate drug discovery and development by leveraging our rapidly evolving, proprietary platform across biology, chemistry, and clinical development.

For more info, visit our website https://insilico.com

About the Role

We are seeking a Machine Learning Scientist with expertise in modern generative modelling and structure-aware machine learning to contribute to the development of advanced AI systems for modelling complex three-dimensional molecular data. In this role, you will develop deep learning approaches that operate on spatial molecular representations, integrate physical and geometric constraints, and support computational workflows for analyzing complex molecular systems.

Place of work

Level 6, Unit 08, Block A, IRENA HQ Building Masdar City, Abu Dhabi United Arab Emirates

Reports to

Head of AI for Chemistry Solutions

Responsibilities

Model Research \& Development

• Develop machine learning models to analyze and model three-dimensional molecular structures and interactions.

• Design computational workflows for evaluating and prioritizing candidate structures based on predicted structural and physicochemical properties.

• Build architectures that integrate multiple predictive tasks across structural modelling and interaction prediction.

• Develop representations and embeddings for complex molecular geometries and spatial relationships.

• Work with large-scale datasets containing structural and coordinate-based molecular information.

Technical Leadership

• Contribute to the design of scalable pipelines for training models on large structural datasets.

• Define modelling approaches that incorporate spatial context, interaction interfaces, and geometric constraints.

• Collaborate with engineers to ensure efficient training, inference, and integration of models into internal platforms.

Research \& Strategy

• Stay up to date with advances in protein design, molecular ML, and geometric deep learning.

• Evaluate emerging methods (All-atom diffusion models, graph networks, multimodal foundation models).

• Contribute to internal research directions and experimentation with new modelling paradigms for complex spatial data.

General Requirements:

I.              Education

PhD or MS in Machine Learning, Computational Biology, Structural Biology, Computer Science, Biophysics, or related field.

II.           Experience and Skills

Technical background

• Strong experience developing machine learning models operating on 3D spatial or geometric data, such as:

o   diffusion-based generative models

o   graph neural networks

o   equivariant neural networks (SE(3)/SO(3))

o   transformer-based architectures applied to structured data

• Proficiency with PyTorch.

Domain Expertise

• Understanding of molecular structure, spatial interactions, and physical constraints in molecular systems.

• Experience working with coordinate-based structural datasets and molecular data formats.

Engineering Skills

• Experience with distributed training, model optimization, and high-performance compute environments.

• Strong software engineering practices (Git, testing, reproducibility).

Preferred Qualifications

• Familiarity with modern machine learning approaches for macromolecular structure modelling and prediction, including systems such as AlphaFold or related frameworks.

• Exposure to emerging structure-aware generative or modelling architectures in scientific machine learning.

• Experience with open-source research systems used in structural modelling pipelines (e.g., Boltz or similar tools).

• Background in computational chemistry, structural biophysics, or molecular simulation.

• Experience developing geometry-aware or constraint-aware machine learning models.

• Contributions to open-source machine learning or scientific computing projects.