Genomic data is moving beyond the laboratory and into the infrastructure of modern health systems. As sequencing becomes cheaper and more accessible, the challenge is no longer generating more data but building platforms that can process, connect, govern, and ultimately turn that information into useful decisions at a population scale.
For Robert Morfino, Business Development Lead at Verily, that shift has made genomics as much a question of infrastructure and policy as technology. “I believe genomic data has become national infrastructure,” he says, arguing that the countries and companies that recognize its strategic value will help shape the next decade of precision health and health security.
Building Platforms That Can Scale
A scalable genomics platform has to make sophisticated infrastructure accessible without sacrificing security or scientific rigor. Morfino points to four principles that can help achieve that balance: cloud computing, curated workflows, multimodal data integration, and federated learning.
Cloud infrastructure can democratize access to high-performance computing and storage, allowing researchers to process enormous datasets without maintaining expensive local hardware. But computing power alone is not enough. “We have to remove the technical friction for the people actually using the platform,” Morfino says.
That means packaging standardized bioinformatics pipelines into intuitive, one-click tools rather than requiring scientists and clinicians to manage code, dependencies, and command-line interfaces themselves. The goal is effectively an app ecosystem for bioinformatics, where reproducible analysis becomes accessible to a much broader group of users.
The platform also needs to connect genomic information with clinical records, laboratory results, and imaging. Finally, as genomic infrastructure expands across jurisdictions, federated learning can support data sovereignty by bringing algorithms to sensitive data rather than moving raw patient sequences across borders.
Making Genomic Data AI-Ready
Simply accumulating genomic data does not make an organization ready for AI. “Most organizations fail at making genomic data AI-ready because they confuse storing massive amounts of sequence data with preparing data for machine learning,” Morfino says.
The first problem is context. A genome alone cannot provide the clinical labels AI models need to identify meaningful patterns. Genomic variants need to be connected to patient outcomes, electronic health records, lifestyle information, and other biological datasets using standardized approaches.
The second is workflow design. Ad hoc scripts, manual transfers, and fragmented pipelines can introduce inconsistencies and technical batch effects. Standardized, containerized, and automated workflows are critical to ensuring that models learn biological signals, rather than artifacts created by the way data was processed.
Governance must also be designed from the beginning. Centralizing genomic information and attempting to address privacy later can create major barriers when organizations expand into cross-border research or clinical collaborations. Privacy-preserving architectures and fine-grained access controls are far easier to build into genomic infrastructure than to retrofit later.
Commercializing Complex Genomic Infrastructure
Technical sophistication does not automatically translate into adoption. Morfino’s experience negotiating major government and private-sector contracts has shaped an approach to commercialization centered on understanding the broader organization, not selling the technology.
Different stakeholders can have entirely different definitions of value:
· A bioinformatician may want to eliminate pipeline bottlenecks, while a chief financial officer may be focused on predictable infrastructure costs.
· A government agency may prioritize data sovereignty.
That makes stakeholder alignment essential. Major contracts often involve executives across research and development, medicine, engineering, IT security, legal, compliance, and procurement. A strong internal champion can open the door, but unresolved concerns elsewhere in the organization can still derail a deal. Morfino describes the process as “a collaborative effort to understand an organization’s need and tailor the solution around it,” while making the purchasing decision as safe and seamless as possible.
Preparing for Genomics as Everyday Infrastructure
The larger shift may be the normalization of genomics itself. Healthcare leaders have often treated sequencing as a specialized tool for rare diseases or advanced oncology. Morfino sees a future in which genomic information becomes a routine component of healthcare, integrated into preventive medicine, primary care, and prescribing.
That transition will create new infrastructure demands. Falling sequencing costs are shifting the economic burden toward compute, data orchestration, and storage. At the same time, advances in long-read sequencing are making comprehensive genomic analysis increasingly practical.
Automation will also be critical. Scaling genomics cannot depend entirely on expanding the pool of highly specialized experts. Automated bioinformatics and clinical AI will increasingly be needed to translate complex biological data into information that frontline clinicians can use.
The implications extend beyond individual healthcare systems. Genomic surveillance and biosecurity programs depend on the ability to generate, analyze, and share information quickly, including across national borders. Yet the rules governing cross-border genomic data, ownership, and sovereignty are still evolving. The organizations building these systems now are determining how genomic data can move from sequencing pipeline to policy decision and, ultimately, how biological information becomes a strategic asset.
Follow Robert Morfino on LinkedIn.
