Adrian Myers

Co-founder and CTO. GEO offers novel technology insurance claims management, focusing on payer rules extraction for pre-submission validation and transformative bottom-up performance analytics. We solve problems through technology rather than manpower, and focus on transparency and responsiveness in our UX and API design. We are committed to deterministic, auditable development (no vibecoding, no third-party blackboxes). We can strengthen and simplify claims logistics, with analytics to show how.

Pi Health is a health technology and clinical research company committed to transforming global access to innovative medicines and clinical trials.

Reprising role as technical lead for FutureScape, a modeling and simulation asset. Having previously developed and secured a patent for ASim, the distributed computing platform underlying the simulations, I am now developing novel platform capabilities and simulation models for new problem spaces, particularly in supply chain and contested logistics.

The Applied Innovation Group within BeiGene identifies novel technologies or workflows that can reduce the costs of oncology clinical trials and drug development, improve the quality of data capture, or both. Focused heavily on role as architect, development lead, and partner liaison for FICS (“Front-end Interoperable Capture System”), a large health data application spanning the roles of an EMR (electronic medical record), EDC (electronic data capture), CTMS (clinical trial management system), CDMS (clinical data management system), and eTMF (electronic trial master file). The goal of FICS is to reduce costs in drug development clinical trials by creating a unified data model and role-based, workflow-forward application that eliminates silos and simplifies many realtime operations. When complete, FICS is planned to be deployed in the US, China, Australia, and India.

• Developed and patented ASim, a custom distributed application framework to power multi-intelligence agent-based simulations, built from the ground up around low-level code with minimal internal or external dependencies. The framework includes a dynamic load balancer, high-performance AI processors, and Unity3D-based visualization. The framework emphasizes high performance (low-contention concurrency, custom serialization, carefully planned memory allocation, and socket operations tuned for low-latency throughput), reproducibility (global clock, deterministic traversal, reproducible randomizers), is cloud-agnostic and is easily ported or extended. ASim has been adapted to run traditional, high-iteration simulations under a reinforcement learning paradigm. Deloitte’s patent application for ASim was approved with all 49 of its original claims in 2020, making it the fourth patent held by Deloitte US firms.

I joined FDA through the DHHS EIR program, where I brought my alliance-building skills and inter-disciplinary data science background to catalyze internal and external cancer data aggregation and analysis projects in OHOP. FDA holds an unparalleled dataset comprising all clinical trial data submitted for US drug approvals, spanning ~3,500 variables of identifiable-patient observational data. I partnered with medical officers on research projects leveraging our unique perspective on the oncology pipeline and building software and tools to aggregate, harmonize, normalize, make interoperable and searchable, and analyze clinical trial data, resulting in several manuscripts and posters and a presentation at ASCO. These included analyses of demographic composition in clinical trials in oncology over a 10-year window, analyses of predictive power of baseline genomic markers for I-O products, and disease-specific uses of comparative statistical tools (e.g., AUC measures of depth and duration of response). I developed machine learning methodology for CDISC (SDTM and ADaM) datasets. I built custom applications for extracting and normalizing data in CDISC format from SAS XPT documents and related metadata and labels, involving extensive structured and unstructured data mining, dictionary normalization, reconstruction of derived fields, and manual curation. I also contributed to custom algorithms for genomics data in the High-performance Integrated Virtual Environment (HIVE) NGS pipeline. Externally, I participated in industry events and public-private partnerships promoting progressive thinking in health data, with topics ranging from real-world evidence generation and research (via a partnership with VA) to incentivizing data-sharing financially through the creation of a Blockchain-driven health data marketplace in collaboration with IBM Watson.

HumanGeo is a unique technology company with a focus on the intersection of big data applications, social media analytics, and geospatial visualization. I came to HumanGeo as part of a small cabal from Ringtail Design that stayed together after Ringtail closed its doors. While Ringtail focused heavily on user experience and graphically-intense bespoke product development, HumanGeo focuses instead on enterprise service development, machine learning, and data science tools which meet specific needs of analysts in the field. The environment at HumanGeo featured linguists, mathematicians, and former operatives working shoulder to shoulder with developers on almost every problem. Our working relationships, and the resulting products, exemplified the value of inter-disciplinary teams focused on specific use-cases. It is difficult to foster a strong, tight-knit culture where one does not exist. Often, the best you can do is to try to preserve one, which is especially difficult when a company more than doubles in size. HumanGeo has one of the strongest cultures I have seen, which it strengthened while making a dramatic jump in size, while raising the already high bar for technical competence across the entire company. Watching the company manage this feat taught me a lot about what is possible with thoughtful leadership.

Ringtail Design is a data software development company that built the most intuitive situational awareness and data visualization tools in the world. Ringtail went out of business abruptly a few months after I started. HumanGeo, a local competitor, generously offered to keep the core of our team intact at their office. Ringtail's flagship product, Replay, is nearly unique among C# 3D visualization applications and its unusual demands forced the original architects to build expert knowledge about the workings of the MS CLR and a wide range of supporting technologies. Topics in an afternoon of coding would range from comparing atmospheric scattering implementations to discussions of ideal data structure growth rates. Our culture of commitment to our users and their mission, combined with a deep understanding of our technology and how to use it most effectively, was very positive, was a common thread running through Ringtail and HumanGeo.