TriNetX Wins Pinnacle Technology Award: The Case for AI Built on Trustworthy Data

Key takeaways:

• TriNetX has been named a Platinum winner in the Pinnacle Technology Awards’ AI Solutions for Clinical Research & Data Insights category, a recognition that reflects the maturation of AI purpose-built for clinical research. 
• Trustworthy AI in this field starts with the data foundation, not the algorithm. Our federated network of 300 million patients across 240+ healthcare organizations is what makes our AI work in practice. 
• From expanding recruitable patient pools by 189% to predicting cancer risk up to 18 months before diagnosis, our AI is producing measurable outcomes, not just capabilities. 

The Pinnacle Awards named TriNetX as a Platinum winner in AI Solutions for Clinical Research & Data Insights at the 2026 Pinnacle Technology Awards, an annual program honoring companies and innovators redefining what is possible across the global technology landscape. 

This year’s program drew nominations from across the technology industry. We were recognized alongside winners in categories spanning enterprise AI deployment, generative AI, cybersecurity, and healthtech, a field that reflects how broadly AI has embedded itself into critical industries. Within that competitive landscape, we stood out for a reason that runs deeper than feature development. 

The Foundation Behind the Recognition

This award reflects a specific argument we have been making and proving for years: AI built on weak foundations crumbles. 

The industry has no shortage of AI tools applied to clinical research. What most of them share is a dependency on data that was not built for this purpose, aggregated from opaque third parties, inconsistently normalized, or static where clinical reality is continuously changing. Feasibility predictions miss. Risk models underperform. Trials still fail to meet enrollment targets at rates that have barely moved in decades. 

We took a different path. Rather than building AI on top of whatever data was available, we built our data infrastructure first. The result is a federated network that sources data directly from more than 240 healthcare organizations across more than 20 countries, spanning over 13,000 clinical sites and 300 million patients. Every data element traces back to its source. The network is continuously refreshed as treatments and care patterns evolve. 

That foundation, comprehensive, high-quality, transparent, and current, is what makes our AI layers actually work. 

AI Capabilities Built for Research Workflows

On top of that data infrastructure, we have developed AI capabilities designed around how clinical research teams work, not around showcasing AI for its own sake. 

A conversational AI interface lets researchers query real-world data (RWD) in plain language. Enhanced API capabilities allow pharmaceutical partners to submit queries and receive patient counts, feasibility metrics, and site intelligence in real time from within the systems they already use. No platform-switching, no custom query development. A human-in-the-loop active-learning pipeline applies named entity recognition and assertion models to unlock clinical information buried in unstructured notes, increasing smoking-status coverage from 22.3% in coded data to 66.9% when notes are included, with confidence-scored and auditable outputs. 

Emerging agentic AI capabilities extend this further, helping teams move from retrieving information to advancing governed workflow steps inside their existing tools. 

What This Looks Like in Practice

The results speak to why this work matters. 

AI-optimized eligibility criteria expanded a UK-Germany COPD trial’s recruitable patient pool by 189%. Precision-targeting models lifted high-risk patient identification from 33% to over 85% for Crohn’s disease and over 70% for ulcerative colitis. In lupus trials, 103 candidates were identified across 11 sites before they formally met criteria. 

Early-detection AI trained on TriNetX data is also reaching patients directly. A pancreatic cancer risk model built on 35,000 cases forecasts risk up to 18 months before diagnosis. The LIRIC model for hepatocellular carcinoma, developed across 46,679 cases from 93 healthcare organizations, predicts risk six to 36 months before diagnosis at an AUC of 0.93 with international validation across the United States, Latin America, and Asia-Pacific. 

What the Industry Is Responding To

This Pinnacle recognition fits a broader pattern we have seen over the past 18 months, from the MedTech Breakthrough Clinical Trial Innovation Award to Best of Show recognition at SCOPE and SCOPE Europe. Taken together, these signals point to something larger than any single award: the industry is reassessing what AI in clinical research should look like, and I believe we are building the answer. 

The answer, increasingly, is not algorithms applied to whatever data is at hand. It is AI grounded in data infrastructure rigorous enough to support the next generation of autonomous, agentic capabilities, not just today’s tools. 

That is where we are building. Winning this award is meaningful, but what matters most to me is what it points toward: a future where every research team, regardless of technical expertise, can move from question to answer to action on data they can trust. That is what it means to be The Global Truth Engine for Better Human Health™. 

About Akiko Shimamura

Akiko is an experienced leader in the life sciences industry, having held a range of senior roles. As Senior Vice President at TriNetX, she is responsible for developing products and managing teams focused on trial design and optimization. As the former Vice President of Medidata, she has overseen products across real-world evidence (RWE), commercial analytics, and tokenization. Akiko has a wealth of experience in consulting having previously worked at McKinsey & Company where she provided advice to companies on life sciences and analytics.