When software meets biology.

One convergence, several different markets

Our thesis is that the meeting of software and biology will produce a broad family of businesses: continuous care tools, performance services, research infrastructure and software for biological engineering. The common thread is a tighter connection between measurement, interpretation and action. Yet treating all of this as one digital-health market is a strategic mistake.

A hospital, a pharmaceutical company, a sports organisation and a longevity clinic buy for different reasons. They operate with different evidence requirements, procurement processes and budgets. A successful market-entry plan must choose which buyer will pay, whose workflow will change and what improvement can be demonstrated. A technically elegant platform without those answers risks becoming another pilot that never reaches routine use.

The evidence that the boundary is moving

The FDA maintains a list of AI-enabled medical devices authorised for the US market. It explicitly describes the list as incomplete. This establishes that clinical AI is already a product category; it does not establish that every AI application is clinically effective or that an authorised device has a profitable business model. FDA device list ↗

At the consumer boundary, the FDA cleared Dexcom’s Stelo as the first over-the-counter continuous glucose monitor in March 2024. Its intended population excludes insulin users and people with problematic hypoglycaemia. That precise scope matters: wider access to measurement is a market signal, not permission to make unrestricted medical claims. FDA Stelo announcement ↗

In research, AlphaFold 3 extends structure prediction to interactions involving proteins and other molecule types. The strategic implication is a larger role for computational tools in the research process. A predicted structure, however, is not a validated medicine; experimental work remains essential. Google DeepMind and Isomorphic Labs ↗

Market one: make continuous care operational

The attractive opportunity is rarely another dashboard. It is a closed workflow that turns relevant information into a timely action: a reviewed exception, a follow-up appointment or a documented handover. Providers need tools that fit staffing capacity and avoid creating an unmanageable stream of alerts. Patients need a service they can understand and continue using.

Potential products include monitoring orchestration, referral coordination, patient onboarding and quality-control software for incoming device data. Revenue could come from provider subscriptions or a per-enrolled-person service fee. Reimbursement is a separate question to test in each market, not an assumption to hide inside a financial model.

The first pilot should compare the proposed workflow with current practice. Measure staff time, completion of follow-up, data completeness and the burden placed on users. Where clinical outcomes are claimed, agree an appropriate study with qualified specialists. A reduction in clicks is operational evidence; it is not evidence of improved health.

Market two: longevity needs an evidence layer

Longevity creates demand for longitudinal records, coordinated testing and understandable interpretation. It also creates a temptation to turn uncertain biomarkers into confident promises. The National Institute on Aging describes research connecting epigenetic age measures with later health outcomes. That supports continued investigation, while leaving a major gap between a research association and proving that a commercial intervention extends healthy life. NIA research summary ↗

For software companies, a credible entry point is the infrastructure around evidence: measurement provenance, laboratory integration, repeat-testing consistency, cohort analysis and transparent reporting. Clinics may also buy scheduling, longitudinal care coordination and consent management. These are practical businesses even if more ambitious biological-age claims take years to mature.

Position the product around decisions it genuinely improves. Explain uncertainty and distinguish measured values from inferred scores. A clear record of what changed, what remains unknown and what requires clinical interpretation can become a stronger differentiator than a visually impressive single-number health score.

Market three: human performance beyond the elite

Sport provides a demanding environment for combining training history, movement data, subjective feedback and recovery information. Our forecast is that useful capabilities will extend into rehabilitation, recreational sport and selected workplace programmes. The opportunity is a better coaching or care workflow, rather than unrestricted surveillance of people.

Promising offers include coach-facing planning tools, adherence support and systems that make multi-device data comparable over time. The buyer could be a club, a rehabilitation provider or an established equipment brand. A B2B partnership can supply distribution and domain knowledge that a new consumer application would have to build at considerable cost.

Keep the intended use precise. A training recommendation, an injury diagnosis and an employment decision have different consequences. Data collected for one purpose should not silently become a score used for another. Test whether insights change a coach’s decision or improve user adherence before adding further sensors and model complexity.

Market four: the software around bioengineering

Biology is becoming more computational, but laboratory work remains full of handovers between instruments, people and data systems. This creates opportunities in experiment tracking, sample traceability, laboratory orchestration, model evaluation and reproducible analysis. Many software firms can enter through these enabling layers without claiming to discover a drug themselves.

The commercial unit is often a research programme or a laboratory workflow. Buyers value saved scientist time, reproducibility, reliable instrument integration and faster identification of failed experiments. A specialised product can start with one instrument family or one experimental method and expand through partnerships.

Remote clinical research is another adjacent opportunity. FDA guidance already addresses the use of digital health technologies to acquire trial data remotely. Companies can build tools for data quality, participant support and device operations, while recognising that suitability depends on the investigation. FDA remote-data guidance ↗

Europe’s infrastructure window

The European Health Data Space entered into force in March 2025, with major provisions scheduled to apply progressively in 2029 and 2031. It creates a long implementation window for interoperability and governed data use. It does not create an immediately available pool of patient data for commercial exploitation. The Commission describes controlled secondary access and prohibits use of that route for marketing. European Commission ↗

Companies should build competence in data provenance, identity, access control and health information exchange. HL7 FHIR provides a standard for exchanging healthcare information, but adopting a standard does not make systems interoperable automatically: local profiles, terminology, data quality and workflow still matter. HL7 FHIR overview ↗

A market-entry programme that can survive procurement

Start with one country, one buyer type and one costly workflow. Interview both the budget owner and the people who must use the product. Map existing systems, contracting constraints and the evidence needed for a purchase. Choose a narrow initial offer with a defined deployment effort and a measurable success criterion.

Build a team that combines engineering with domain judgement: a clinical or scientific partner, an interoperability lead, a product owner and access to regulatory and quality expertise. For performance products, include sport-science competence. For bioengineering, include someone who has run the laboratory workflow being automated. Hire for the specific market rather than a generic health-tech label.

Commercial communication should connect the technical capability to an evidenced operational benefit. Publish study design, scope and limitations; equip partners with accurate claims and implementation materials. Treat clinician trust, laboratory relationships and integration partners as distribution assets. The firms most likely to win will make advanced technology usable inside a real decision process—and give a clearly identified buyer a reason to fund it.

A practical financial model should separate implementation revenue from recurring software revenue and include the cost of integrations, evidence generation, support and procurement time. Test a paid pilot with an agreed route to a larger contract. An illustrative success condition might combine an operational improvement with sustained usage and a named budget for rollout; the exact thresholds should be set with the buyer. This discipline helps distinguish an interesting scientific collaboration from a repeatable business, while leaving room for both to create value.

Sources & further reading

Primary sources consulted for this essay. Links and public announcements checked during preparation on 20 October 2026.

1. FDA — AI-enabled medical device list ↗
2. FDA — First over-the-counter continuous glucose monitor ↗
3. Google DeepMind and Isomorphic Labs — AlphaFold 3 ↗
4. National Institute on Aging — Epigenetic age research ↗
5. FDA — Digital health technologies for remote data acquisition ↗
6. European Commission — EHDS regulation and phased implementation ↗
7. HL7 — FHIR overview ↗