Jason Pearson hopes that chemists around the world will soon be using a machine-learning system to accelerate the discovery and development of new compounds.

In fact, he’s preparing to release such a platform to a restricted group in July.

A professor at University of Prince Edward Island, Pearson is a computational chemist — that is, a chemist who uses computers in the development and analysis of compounds. He’s also the head of Retrievium Inc., the Charlotte-town-based startup that is developing the platform.

For the past three years, Pearson has been collaborating with Ray Poirier, a chemist at Memorial University of Newfoundland, and with another Charlottetown startup, discoverygarden, to build the platform.

They’ve produced algorithms that recognize molecular patterns and properties and automatically lead researchers to new compounds that might have similar properties.

Discoverygarden, which oversees a digital repository for archives and museums around the world, worked with the team on developing the dashboard.

“If we compile a big enough data bank for chemists, then the computers we have get will smarter and smarter and smarter and eventually they’ll do a lot of things that humans can’t do,” said Pearson in an interview on Monday.

He explained that elements can be combined in an infinite number of groupings, and chemists in industry and research institutes spend their lives in search of better compounds than previously existed.

Charlottetown's RevIQ Plans to Double Staff

Retrievium will allow chemists in a variety of fields to input their data on compounds, the elements used and the properties of the molecules.

As the databank grows, the algorithms can make intelligent decisions on how to group together different elements to produce better drugs or industrial materials.

It would analyze the composition of chemicals and how different chemicals interact with one another.

“We’re putting together a business model,” said Pearson. “We’ve got some anticipated revenue streams that we’re going to have to test. We’re looking at industrial clients that have a lot of data and need a way of managing that data. It includes Big Pharma, of course, but it also includes industrial companies.”

He added Retrievium is also considering marketing the platform to data centres and academic researchers, who could benefit from its artificial intelligence.

The company will release a working prototype to about a dozen close collaborators in July, and then seek feedback from them.

Depending on the response from these working chemists, Pearson is hoping for a general release late this year or early in 2017.

Pearson and Poirier have fund-ed the project so far from various sources, largely from Springboard Atlantic, the organization that helps to commercialize Atlantic Canadian university research, and NSERC, which helps companies work with academic researchers. Pearson said their universities, especially the UPEI tech transfer office Synapse, helped as well.

In the future, they plan to seek private funding for the company and would be especially interested in teaming up with a corporation that would use the platform.

“If there’s an industrial client who sees a lot of potential, there’s a chance they will underwrite the development,” he said. “As researchers who developed the tool, we really want to see this tool grow.”