On June 8th, 2026 OmniScience CEO Angela Holmes will take the stage at the Digital Health and AI Innovation Summit, hosted by World BI Group in Boston. The outline below previews her presentation. Visit DHAI to learn more.
Most organizations still operate in a world of delayed data reconciliation, fragmented systems, spreadsheet-driven workflows, and retrospective oversight. Teams wait days - or weeks - for signals to emerge, issues to surface, and decisions to be made. By the time insights reach the right stakeholders, opportunities to intervene may already be lost.
But the model is changing.
Across the industry, we are seeing the early foundations of a new paradigm: real-time clinical development. Regulators are signaling it. Sponsors are demanding it. AI and data infrastructure are finally making it possible.
The question is no longer whether real-time trials are coming.
The question is whether pharma is operationally prepared for them.
Recently, the FDA outlined a new operating model for clinical development that reflects a broader evolution in how therapies will be evaluated and monitored. The agency is emphasizing more continuous data review, modernized evidence generation, adaptive approaches, and technology-enabled oversight.
This is an important signal to the industry.
The future of clinical development will not be built around static milestones and periodic reporting cycles. It will increasingly depend on continuously updated, analysis-ready data and the ability to generate explainable insights in near real time.
That changes everything about how trials need to operate.
Real-time engagement with regulators requires more than dashboards. It requires:
Many organizations are still trying to accomplish this with disconnected systems and manual processes.
That gap is becoming impossible to ignore.
One of the biggest barriers to real-time trials is not the science.
It is the operating model.
Clinical development teams today work across dozens of siloed systems: EDC, CTMS, labs, safety databases, ePRO, wearables, spreadsheets, emails, and slide decks. Data lives everywhere. Context lives nowhere.
As a result:
The irony is that we have more clinical data than ever before - but less operational clarity.
The industry does not need more data.
It needs a way to transform fragmented data into real-time knowledge.
This is where AI-native infrastructure becomes critical.
At OmniScience, we believe the future of clinical development is a continuously intelligent operating environment - what we call a clinical trial control tower.
That vision led us to build Vivo, the first agentic AI-powered control tower purpose-built for clinical trials. Vivo unifies fragmented clinical, safety, and operational data and delivers real-time, explainable insights that help teams move from reactive reporting to proactive oversight.
Importantly, this is not about replacing human expertise.
It is about amplifying clinical teams’ ability to see, understand, and act faster.
When teams can interact with live, unified trial data in real time, they can:
This is how trials become adaptive, responsive, and continuously informed.
And increasingly, this is what regulators will expect.
One of the biggest misconceptions in clinical AI is that integrating data sources is enough.
It is not.
Most organizations already have data warehouses, dashboards, and reporting layers. But simply joining together CTMS, EDC, lab, and safety data does not create operational intelligence. It creates a larger pool of disconnected information.
To operate clinical trials in real time, systems need to understand the meaning of what they are seeing.
That is why Vivo was built around a clinical trial workflow ontology.
At its core, an ontology is a structured, machine-readable model of a domain. In clinical development, that means encoding what a clinical trial actually is: how protocols are structured, how amendments affect downstream workflows, how sites relate to patients, how endpoints connect to statistical analysis plans, and how operational events impact regulatory timelines.
This becomes critically important in real-world trial operations.
For example, a patient dropout at Site 12 during a specific enrollment period is not just a data point. It may be operationally meaningful because the protocol requires a certain retention threshold to maintain statistical power for the primary endpoint.
Similarly, a visit window deviation captured in EDC is not merely an isolated event. In context, it may have implications for protocol deviations, statistical analysis, database lock timelines, or ultimately the clinical study report.
Without a semantic understanding of these relationships, AI systems can only surface anomalies.
With an ontology, they can reason about significance.
That semantic layer is what allows Vivo’s AI agents to generate explainable, protocol-aware recommendations rather than simply producing statistical alerts or disconnected summaries.
In other words, Vivo is not just connecting systems.
It encodes the operational logic, workflow dependencies, and regulatory context of clinical development itself.
That distinction matters enormously as the industry moves toward real-time trials.
We are already seeing this shift happen in practice.
In our work with sponsors and partners, the value of real-time intelligence becomes immediately clear once fragmented data is unified. Teams move from searching for answers to acting on insights.
For example, Vivo has been used to help unify data across EDC, CTMS, labs, safety systems, and patient-reported outcomes - allowing clinical teams to identify discrepancies, surface operational risks, and answer critical questions while patients are still in clinic.
These are not incremental improvements.
They represent a fundamentally different way of operating clinical trials.
There is also an important misconception emerging in the industry: that generative AI alone will solve clinical trial inefficiencies.
It will not.
Real-time trials require trustworthy AI grounded in validated clinical data, explainability, governance, and operational context.
In highly regulated environments, AI cannot function as a black box.
Clinical teams and regulators need to understand:
This is why explainability, provenance, and human oversight are essential design principles for the next generation of clinical AI systems.
The winners in this next era will not simply be the organizations using AI.
They will be the organizations using AI responsibly, transparently, and operationally.
Some organizations are.
Many are not.
The companies that will lead the next decade of clinical development are already investing in:
Others remain constrained by legacy infrastructure and fragmented operating models.
The shift to real-time trials will not happen overnight. But the direction is now unmistakable.
Clinical development is moving from:
This is not simply a technology transition.
It is an operational transformation.
And the organizations that embrace it early will gain a significant advantage - in speed, efficiency, quality, and ultimately patient impact.
Real-time trials are coming.
The industry now has to decide whether it is ready to operate in real time too.
