Sovereign AI invests in Potential Sciences to improve the data behind molecular AI

We are announcing our investment in Potential Sciences, a UK startup that has today emerged from stealth to develop technology using AI and high-performance computing to better predict how molecules behave.

By simulating chemistry from the laws of physics, Potential Sciences aim to generate highly accurate scientific data at a scale previously too expensive to achieve. This will help AI models better predict molecular behaviour and support advances in drug discovery, materials science, and more.

Why chemistry needs better data

AI models can help researchers identify promising molecules, predict how they interact with potential drug targets, and narrow down which compounds to investigate. But their predictions often depend on the quality of the data they learn from.

Experimental chemistry data is expensive and time-consuming to produce. Existing computational datasets can fill some of that gap, but many rely on approximate calculations that struggle to capture the small differences that can determine whether a drug binds to its target or a chemical reaction works as expected.

More accurate methods exist, but the computing power traditionally needed to run them on CPUs has made it difficult to generate data at the scale modern AI models require. Potential Sciences is tackling this problem by rebuilding quantum chemistry software to run efficiently on GPUs, which can perform these calculations at a much lower cost. Its technology runs at a reported 100 times lower cost than traditional CPU-based implementations.

Potential Sciences will use this data to develop models that learn how atoms interact, with the aim of helping scientists predict molecular behaviour before committing time and money to laboratory experiments. Sovereign AI is supporting this work with access to compute through the AI Research Resource (AIRR). This compute will help Potential Sciences run its technology at scale and generate millions of highly accurate chemistry calculations for AI models to learn from.

The initial focus for Potential Sciences is on small-molecule drug discovery. By better predicting how molecules behave, its technology could help researchers identify which potential drugs are worth testing and how well they might work. The same underlying chemistry could also support research into catalysts, enzymes, materials and more.

Hatem Helal, Founder and CEO of Potential Sciences:

"AI for drug discovery has learned almost everything the existing experimental record can teach it. The next leap needs new data, and the laws of physics can supply it, as long as your software can use modern hardware.

We rebuilt quantum chemistry for GPUs so that gold-standard accuracy is affordable at the scale AI needs. When your code is 100 times more efficient, a seed round buys what used to take a dedicated datacenter."

Josephine Kant, Investment Partner at Sovereign AI:

“I first met Hatem in Cambridge, and he immediately struck me as a formidable founder. He combines technical excellence with low ego and unapologetic ambition. He knows the science in extraordinary detail, but he’s also deliberate about bringing in people who are exceptional in areas different from his own.

Potential Sciences have the ambition to become a true frontier lab for chemistry, with the potential to fundamentally change how the field works.”