The building blocks of AI: 15 startups show the future of biotech innovation

The convergence of AI and the healthcare sector is sparking a new era of innovation, with startups leading numerous advancements in medicine, care, and research.

There’s increasingly popular clinical tools to manage routine tasks and time-consuming record-keeping, as well as the many notable efforts at novel drug discovery using large language models. There’s also more niche and unexpected applications underway, like a mobile sonogram device that can take breast cancer screenings into the field, or “bioprinted tissue” that can support, repair, or even replace certain biological functions in the body.

These are just a few examples of the cutting-edge concepts being pioneered by healthcare and life science startups building on Google Cloud. It not only takes considerable industry expertise but also the right AI tools and techniques to unlock projects of this scale and importance. That’s why medtech startups are increasingly turning to Google Cloud, to access purpose-built AI architecture and industry-specific models and technologies, such as GoogleDeepmind’s AlphaFold models and Health AI Developer Foundations, a collection of Gemma-based open models that we hope will help improve the efficiency of AI-powered drug discovery and help developers build AI models for healthcare.

Healthcare and life-science startups know how critical it is to have the right tools and knowledge if they are going to accelerate research, advance biological understanding, transform drug development, and personalize medicine. Of all the industries currently being transformed by AI, healthcare and biotech has shown some of the fastest and most dramatic gains, which is why we are so pleased to support these efforts through our Google for Startups Cloud program.

Already, hundreds of organizations are looking for the latest cures for diseases, ways to ease the burdens on stretched clinicians, develop new biomedical tools and devices, and tackle some of the oldest and toughest problems in medicine. We wanted to share the work of more than a dozen leading the charge, and hope that you will see some possible inspiration or collaboration between their work and your own.

Like a good medical team at the hospital, we are bound to do our best work building together, on each other’s strengths.

Unlocking new frontiers of research with AI

In many ways, AI and modern IT infrastructure are making it easier than ever for innovative thinkers to bring their creations from idea to application in an affordable and efficient way. Biotech startups are seizing on the opportunity and helping to improve virtually every element of healthcare, from genomics to public health management.

For example, several biotech startups building on Google Cloud are powering a surge in advanced medical and pharmaceutical research.

• Menten AI is using AI to accelerate the development of peptide therapeutics, using Google Cloud’s high-performance computing and machine learning capabilities to rapidly design and optimize novel drug candidates.
• Via Scientific and Google Cloud are partnering to deliver Via Foundry, an enterprise-grade bioinformatics platform that helps make the drug discovery process more efficient through unified multi-omics data and AI-powered analytics. Researchers and teams using Via Foundry will benefit from Google Cloud’s scalable infrastructure, accessing gen AI capabilities including Gemini and Vertex AI to accelerate discoveries and transform complex biological data into actionable insights.
• Triplebar is building an AI model for genomes for therapeutic production, in conjunction with a proprietary miniaturization process that can test billions of cellular mutations at once, greatly scaling up from current methods that typically only handle a few thousand cells at a time. Triplebar’s goal is to create data sets of sufficient size to train generative AI at the scale of the genome.Such a model could capture both local sequence motifs and long-range dependencies that are critical for understanding protein expression and regulation and can help inform new treatments.
• GenBio AI is on the front lines of developing what it calls the world’s first AI-driven digital organism, or AIDO. Founded by prominent researchers in machine learning and computational biology, GenBio AI’s global operations use Google Cloud to help power their six multiscale foundation models. These models can simulate biological programming to address critical challenges in medicine and biology.
• Moonwalk Bio is a preclinical-stage biotechnology company pioneering new insights and medicines utilizing epigenetic biology and AI, with an initial focus in obesity and cardiometabolic disease. The company leverages its biology expertise and deep computational capabilities to determine causal relationships between genes and their roles in biological pathways that could be therapeutically targeted.
• Congruence Therapeutics is a computationally driven biotechnology company building a unique pipeline of small molecule correctors rationally designed to rescue aberrant protein function. The company's proprietary scalable platform, Revenir, captures the biophysical features of proteins across their conformational ensembles, in order to identify novel allosteric and cryptic pockets which are virtually screened to generate novel chemical matter.
• DNAstack, a leading genomics data management and analysis platform, partnered with Google Cloud to leverage its scalable infrastructure and advanced analytics tools, enabling DNAstack to accelerate research and discovery in personalized medicine.

AI helps us care for patients and tackle problems like never before

Other biotech startups are focusing on empowering clinicians and other professionals who assist patients at the point of care.

• Digital Diagnostics works with Google Cloud to enhance the ability for LumineticsCore (previously IDx-DR) to reach more patients living with diabetes who are most in need of specialty eye care. By leveraging Google Cloud’s secure infrastructure, Digital Diagnostics will protect sensitive health data within national borders, safeguarding patient privacy, maintaining regulatory compliance, and fostering trust.
• iSono Health has pioneered a unique Virtual Sonographer, an intelligent, automated 3D Ultrasound platform powered by Google Cloud AI infrastructure. The platform is designed to bring breast imaging directly to the point of care and empower women and their physicians with fast, accessible, and repeatable imaging – anytime, anywhere.
• Think Research, a leading provider of knowledge-based digital health software solutions, partnered with Google Cloud to use its scalable infrastructure and advanced analytics tools, enabling it to deliver more efficient patient care and improve health outcomes.
• Atropos Health’s GENEVA OS (Generative Evidence Acceleration Operating System) is optimized to work with Google Cloud’s healthcare offerings, including Google Cloud’s Healthcare Data Engine (HDE), leveraging HDE application programming interfaces and BigQuery. This powerful pairing enables customers to efficiently and securely convert data into valuable insights and evidence.
• Aspect Biosystems is a Canadian biotechnology company that develops bioprinted tissue therapeutics. Bioprinted tissue therapeutics are designed to replace, repair, or supplement biological function in the body. These implantable tissues have the potential to transform how we treat some of the most elusive diseases.
• Sully.ai has built an app store for AI agents designed specifically for healthcare professionals, providing support to clinicians on administrative tasks, so they can focus on patients.
• Virgo Surgical provides a leading video capture and management solution for endoscopic medical procedures. To date, the company has captured more than two million high-definition procedure videos — believed to be the largest dataset of its kind. Leveraging this data, Virgo recently announced an AI foundation model for endoscopy, called EndoDINO, which achieves state-of-the-art performance in a wide range of AI benchmarks for endoscopy. Virgo hosts more than 1.75 petabytes of video data in Google Cloud Storage (GCS) and uses Google Kubernetes Engine (GKE) to flexibly scale the ingestion and processing of endoscopy video data.
• Amaros is democratizing ophthalmic care through its holistic and data-driven EvidenceEngine platform that leverages computer vision, machine learning, and AI foundation models to deliver predictive analytics across multiple ophthalmic diseases and indications. Amaros maintains a rigorous ophthalmic database, integrating electronic health and medical records (EHR/EMR), physician notes, multimodal imaging, claims data, and genomics. By connecting previously siloed data sources, Amaros enables healthcare providers to make more informed clinical decisions and life sciences companies to identify new treatment pathways.

Opening image created using Imagen 3 in AI Studio, using the prompt: I'm working on an article about biotech startups building with generative AI. I'd like to make a colorful, upbeat, flat-style illustration that shows a range of biomedical disciplines or practices, including some of these concepts (this is copy directly from the article): "There’s increasingly popular clinical tools built with AI to manage routine tasks and time-consuming record-keeping, as well as the many notable efforts at novel drug discovery using large language models. There’s also more niche and unexpected applications of AI underway, like a mobile sonogram device that can take breast cancer screenings into the field, or a startup building a unique pipeline of small molecule correctors rationally designed to rescue aberrant protein function."

Can you blend and overlay some of the elements, so it doesn't look like each one is so distinct, like these are all elements of a unified whole coming out of the same technology platform?