Solutions

The decisions that move a brand.

SHAARPEC applies trajectory intelligence to three commercial problems: finding patients before they're visible, protecting patients before they're lost, and correcting dosing before outcomes degrade.

Three Applications

One platform. Three decisions.

Each application uses the same trajectory architecture and foundation model — targeted to a specific decision that determines brand performance.

01

Predicting Discontinuation

Which patients on therapy are about to stop?

Refill gaps don't appear in isolation. They're preceded by care fragmentation, urgent care visits, and access friction events that appear in claims weeks before the first missed fill. SHAARPEC identifies this pattern and ranks the HCPs whose panels carry the highest concentration of at-risk patients.

Early pattern detection — identifying at-risk patients before the first missed fill.
02

Detecting Competitive Switching

Which patients are approaching a switch — and which HCPs are losing them?

Flat dose lines, rescue medication onset, shortened refill intervals — these signals are in your claims. SHAARPEC assembles them into a pre-switch trajectory and flags patients before the prescription moves, with a competitive flow map showing which prescribers are losing share and at what velocity.

Pre-switch signal — flagging patients before the prescription moves, not after.
03

Correcting Stalled Titration

Which patients are stuck at sub-therapeutic doses — and why?

Claims contain the titration schedule. Trajectory analysis reads the velocity and pattern of that schedule — distinguishing on-label titration from plateaus and dose regression, and separating institutional under-titration from patient-specific hesitation so the right intervention reaches the right HCP.

Early stall detection — identifying under-titration patterns before outcomes degrade.
The Latency Gap

Your analytics arrive after the decision has already been made.

Conventional claims analytics report on what happened. By the time a quarterly utilization report reaches your field team, the patient has already initiated — or switched — or dropped off. The intervention window is closed.

SHAARPEC reads the same claims differently. It sequences them into longitudinal patient trajectories and trains a foundation model to recognize the patterns that reliably precede clinical events — before those events appear in prescription data.

Scientific Foundation

Methodology validated before it was applied commercially.

Cardiovascular · BMJ Open, 2024

CONSIDERING-AF

Bristol Myers Squibb · Pfizer · Massachusetts General Hospital

AI-enriched trajectory analysis applied to population-scale atrial fibrillation screening across 330,000 patients. The model identified patients at optimal screening windows based on longitudinal care progression — not static risk factors.

5.4× improvement in AF detection. AUC 0.79–0.84, validated internationally.
Published: BMJ Open, 2024
Cardiology · European Heart Journal · JMIR, 2023

HaRP — Heart Failure Readmission

Novartis · Region Halland · Cambio Healthcare Systems

Explainable machine learning trained on complete longitudinal heart failure patient histories. Deployed live inside a commercial EHR system as clinical decision support — the same architecture that underlies SHAARPEC's commercial trajectory models.

€1,130 cost reduction per patient. Model performance matched deep learning.
Published: European Heart Journal · JMIR, 2023
The SHAARPEC platform is built on a methodology with multiple peer-reviewed validations across disease areas and health systems. The two studies above are particularly representative for pharma commercial applications — trajectory-based pattern recognition applied to longitudinal patient data at scale. The trajectory-completion capability builds on published research in graph-native synthetic patient generation — including work recognized in Nature npj Digital Medicine (2023) as the highest-fidelity method for healthcare synthetic data to date.
Current Validation

Being validated right now, with real data.

SHAARPEC is currently running a structured pilot with PrecisionAQ — one of the leading pharma commercialization analytics organizations in the US. A national breast cancer cohort of 23,000+ patients, structured from open and closed claims.

Ready to see this applied to your indication?

We start with your data question and show you what the signal looks like for your therapy.

Request a Demo