Computational Chemist; gn Chemistry AI Venture

This an Atlantic portfolio venture.

This early-stage venture is building a discovery engine for chemistry, comprising a platform for running end-to-end simulations of chemical reaction systems, and a world model + agent stack that uses the simulation platform to discover new chemistry.

Their goal is to build backend scientific data infrastructure to meet the demands of AI systems that can truly accelerate and automate the discovery layer of chemistry.
Both founders hold PhDs in Chemistry from the University of Oxford, and the venture has already secured pre-seed funding. They are now building out their founding team.

• As a Computational Chemist, you will play a central role in shaping the scientific foundations of the venture's technology, helping develop the computational methods and workflows that underpin its end-to-end chemical reaction simulation platform.

• You’ll work closely with researchers and engineers to integrate state-of-the-art computational chemistry methods with machine learning approaches, driving progress towards the venture's strategic and scientific objectives.

• You’ll take ownership of developing, validating and optimising the computational chemistry components of the platform, with a particular focus on reaction mechanisms, reaction kinetics, molecular dynamics and the modelling of complex chemical environments.

• You will design, implement and refine scalable scientific workflows for generating, processing and analysing chemical reaction data, improving the accuracy, reliability and computational efficiency of simulations.

• Beyond your core scientific responsibilities, the team is looking for someone who can contribute to strategic research decisions, bring strong independent scientific judgement, and help shape the interdisciplinary research culture of the company at an early stage.

• Foundation models for chemical reactivity: The team is building and training foundation models for chemical reactivity with a variety of data modalities. You'll help curate this foundational data corpus by developing cheminformatics and computational chemistry workflows for generating data from combinatorial enumeration, as well as robust data filtering, clustering and validation routines.

• Computational methods: The team is working on efficient computational methods from which foundational data is generated and the platform is validated. You'll be responsible for developing protocols for estimating solvation and entropy factors in chemical reaction steps. This involves keeping up to date with published literature for ideas and inspiration.

• Generative models: The venture is expanding its foundational data corpus beyond combinatorial workflows. You'll help design, validate and benchmark the generative model architectures that accomplish this.

• Chemical reaction simulation: The team is building out an end-to-end chemical reaction simulation platform. You'll be involved in designing the architecture and assembling the platform, as well as developing benchmarks and validation protocols against computational and experimental methods.

• World models and agents: The venture will be developing an agent + world model stack that operates on the simulation platform. You'll be responsible for formulating the action space in which the world model is trained and the agent explores new chemical reaction space.

The team is looking for someone who enjoys being given a blank canvas and the chance to create something spectacular with it. In particular, someone who brings:

• A PhD in Computational Chemistry, Machine Learning, or a related field.

• Intellectual curiosity:
  Someone who enjoys venturing outside their comfort zone, exploring new ideas and applying…