If startup funding rounds are any indication, generative AI is well-adopted in science. That's natural. Research and development involves a lot of trial and error, and any tool that can speed up the process for researchers is sure to come in handy.
The latest is Albert Invent, which offers an AI platform trained on data from past chemical experiments, allowing chemists to interrogate molecular-level data when formulating chemicals. The company's founders previously ran a 3D printing company and decided to use their materials science knowledge to train AI on chemical processes. The startup has now secured a $22.5 million Series A funding round led by Coatue.
The company's platform, called Albert Breakthrough, combines structured data with the company's proprietary AI models. We hope that chemical companies will be able to use this platform to develop new and better products faster. The company says its platform can, for example, generate real-time toxicity prediction chemicals and “outperform” standard industry models.
Albert Invent's customers include Chemours, Solenis, Keystone Industries, Applied Molecules, Henkel, and Nouryon.
Nick Talken, CEO and co-founder of Albert Invent, believes the platform will bring to chemical science what data scientists have traditionally had access to. “This is a SaaS product that the world's largest chemical company is using to fundamentally reinvent the physical world. “All of the biggest problems that we have are going to be solved by chemistry,” he said.
Talken said he built a unique foundational model trained on more than 15 million chemical structures. “In this industry, we don't want to just have a corpus of data on the Internet. We need to have domain-specific knowledge. That's why we're using almost every public information space for chemistry, about 15 million He takes molecules and builds fundamental models that are the driving force behind Albert's breakthroughs.”
Asked whether the company uses foundational models such as OpenAI, he said that the company sometimes uses them for part of its agent networks, such as chatbots, but that the basic chemical models are proprietary. He said it was a thing.
His co-founder, Ken Kisner, previously led Molecule, which he started out of a chemistry lab built in a trailer in his backyard. The two founded Molecule Corp to become a global manufacturer of 3D materials for optical lithography, which they sold to Henkel Corporation in May 2019.
And during their time at Henkel, they built a team to tackle the problems they are currently working on. “We basically grew Albert Invent as a software startup within this 145-year-old multinational chemical company and then spun it out on its own,” Token said.
Prior to this Series A round, the company had raised a smaller seed round led by Index Ventures in late 2022. TCV, Index Ventures, F-Prime, and Homebrew also participated in the Series A.
“We are excited to support Albert as they seek to transform the way chemical research is conducted by implementing the latest AI technologies to increase efficiency and realize overall business benefits. ,” Cotu General Partner David Schneider said in a statement.
Johan Landfors, Nouryon's chief technology officer, said in a statement that the platform is now essential to the company's product development.