Apheris: Federated Computing to Unlock AI in Life Sciences

Artificial Intelligence (AI) is transforming industries across the world. But in life sciences, it still faces a major obstacle: sensitive patient data. Health data is key to training AI models, but privacy concerns, strict regulations, and intellectual property issues often prevent its use. Apheris, a German startup, is changing that with its solution: federated computing. This technology allows AI models to train on data without moving it, ensuring both privacy and security.

The Data Bottleneck in Healthcare

AI relies on vast amounts of data to make predictions. However, in healthcare, most data remains locked away. Strict privacy rules prevent organizations from sharing it. Robin Röhm, the CEO and founder of Apheris, explains that these challenges are preventing the growth of AI solutions in life sciences. With data out of reach, it’s difficult for AI to drive advancements in healthcare.

Federated Computing: The Secure Solution

Apheris offers a solution: federated computing. This approach lets organizations run computations locally, where the data is stored. Instead of moving sensitive data, only the results—like model parameters—are shared. This keeps the data safe and private. Federated computing solves the problem of sharing data while protecting patient privacy.

Marcin Hejka from OTB Ventures highlights that federated computing ensures data stays where it belongs: at its source. This reduces privacy risks and makes collaboration between organizations possible.

Apheris and the Growing Federated Data Networks

The future of healthcare AI lies in federated data networks. Apheris is playing a big role in this shift. The company’s platform also works with advanced privacy-enhancing technologies like homomorphic encryption and differential privacy. These technologies add an extra layer of security, ensuring data remains private while still being used to power AI models.

Apheris’ Pivot to Pharma and Life Sciences

At first, Apheris focused on building a federated learning framework to compete with open-source projects. But in 2023, the company shifted its focus. They began targeting the data owners in the pharmaceutical and life sciences industries. This change led to big growth. In fact, Apheris’ revenue increased by four times after launching a new product in late 2023.

The Apheris Compute Gateway: Enabling AI-Driven Drug Discovery

Apheris has developed the Apheris Compute Gateway, a tool that connects local data with AI models in a secure way. This technology is already being used by the AI Structural Biology (AISB) Consortium, which includes companies like AbbVie, Boehringer Ingelheim, Johnson & Johnson, and Sanofi. The consortium is working together to use AI for drug discovery while keeping data private and secure.

Unlocking AI’s Potential in Life Sciences

Apheris plans to focus on more advanced areas like protein complex prediction. These areas have little public data, but a lot of valuable private data. Apheris helps companies access this hidden data while keeping it safe. With their secure platform, life sciences companies can finally tap into the valuable data that was once out of reach.

As Robin Röhm says, “To unlock AI’s full potential, we must address data owners’ concerns about sharing their data.” Apheris is working to do just that—helping AI make a real impact in healthcare.

Conclusion: Apheris Is Shaping the Future of Healthcare AI

Apheris’ federated computing solution is changing how AI is used in healthcare. By allowing AI models to be trained on decentralized data, Apheris ensures privacy and security. The company is paving the way for a future where AI can safely revolutionize healthcare.