Mantis Biotech is creating a ‘digital twin’ for people to help solve the problem of data access

Large-scale languages ​​trained on large datasets can accelerate genomics research, streamline medical records, improve real-time diagnostics, support clinical decision-making, accelerate drug discovery, and generate information for trial design.

But their promises to revolutionize medical research are often in trouble: beyond what conventional medicine relies on, these species struggle in rare diseases and rare conditions, where reliable, representative models are lacking.

From New York Mantis Biotech is said to be developing a solution to the gap in data availability. The company’s platform integrates different data sources to create integrated data that can be used to create so-called “digital twins” of the human body: based on physics, predictive models of anatomy, physiology, and behavior.

The company is putting these digital twins to use in data integration and analysis. These digital twins can be used to study and test new medical procedures, train surgical robots, simulate and predict medical conditions or procedures. For example, a sports team can predict the likelihood of an NFL player developing an Achilles heel injury based on their recent activities, training, nutrition, and the time they have been working, Mantis founder and CEO Georgia Witchel explained to TechCrunch in a recent interview.

To create these twins, the Mantis platform first takes information from various sources such as books, motion picture cameras, biometric sensors, training logs and medical imaging. Then, it uses an LLM-based system to process, validate, and integrate different data streams, and runs all that information through a physics engine to create more reliable interpretations of the dataset, which can then be used to train predictive models.

“We’re able to take all these disparate sources and turn them into predictive models of human behavior. So any time you want to predict human behavior, that’s a great way to use our technology,” Witchel said.

The physics engine layer is very important here, Witchel told TechCrunch, because it helps the platform improve the existing knowledge by implementing the generated content and modeling the physics.

“If I were to ask you to do a hand simulation of a person who is missing a finger, it would be very difficult, because there is no publicly available data about the position of the handwriting of a person who is missing a finger. We can create that data really, really easily, because we just take our physics model and say, remove finger X, change the model,” he said.

Since the Mantis platform fills gaps in information, Witchel thinks it has the potential to be widely used in the healthcare industry, where information about procedures or patients is hard to find, not created or stored in different places. He emphasized serious or rare diseases, where data is difficult to obtain because there are often ethical and regulatory barriers around including patient information in government datasets, or using it to train AI models.

“How do you know when you see a three-year-old running around, and they have a Barbie, and they’re holding it with one leg and smashing it on the table? I want people to have that feeling with our digital twins,” he said. “I think that this will open people to the idea that people can be tested when you are using real people. I feel that right now, people are working with different ideas, which makes sense, because people’s privacy should be respected. In fact, I don’t think that people’s data should be used, especially when you have digital twins.”

So far, Mantis has seen success in elite sports, perhaps because of the need to emulate successful athletes. Witchel said one of the first customers is an NBA team.

“We create digital displays of athletes, where they show how the athlete jumped, not just today, but every day for the last year, and this is how their jumping changes over time compared to how much they are sleeping, or compared to how often they raise their arms above their head,” he said.

The startup recently raised $7.4 million in seed funding led by Decibel VC, with the participation of Y Combinator, a few angel investors, and Liquid 2. The money will be used for recruitment, advertising, marketing and going to market.

The next step for Mantis, Witchel said, is to continue to develop the technology, and release the platform to the general public, to pursue preventive health care. The company is also working to help clinical labs and researchers working on FDA trials, with the goal of providing information about how patients are responding to treatment.