Discovery Lens
C Combination Innovation
Two separate worlds finally connect — and the intersection is a product
The FDA upgraded its shortage database to near-real-time updates in 2024, but clinical systems — EHRs, pharmacy management software — do not connect to it, creating a data availability gap that AI can close in milliseconds where current workflows take hours and often fail entirely.
One-Liner
An AI-powered real-time drug shortage alternative finder that matches unavailable drugs to therapeutically equivalent available alternatives, checking local pharmacy stock and insurance coverage simultaneously.
The Journey
◆Origin
Drug shortages have moved from an occasional inconvenience to a systemic crisis. The FDA active shortage list now regularly exceeds 300 items, covering everything from chemotherapy agents to ADHD medications to GLP-1 drugs. The clinical impact is direct and measurable — patients miss doses, clinicians switch to suboptimal alternatives without systematic guidance, and pharmacists spend hours on manual substitution research that structured data and AI reasoning could handle in seconds.
⚡The Breakthrough
Real-time FDA shortage data meets AI clinical equivalency reasoning combined with live pharmacy inventory APIs. The FDA Orange Book and clinical guidelines for therapeutic substitution are structured, rule-based systems — exactly where AI reasoning performs reliably. The breakthrough produces instant, insurance-aware, clinically validated alternatives at the point of care rather than at the end of a multi-hour manual search.
☠Almost Killed
Regulatory liability for substitution recommendations nearly killed this idea — in most US states, pharmacists cannot substitute a drug without explicit prescriber authorization, and any tool that appeared to recommend substitution autonomously would face immediate legal and compliance pushback. The idea survives as a decision-support tool rather than an autonomous substitution engine: it generates a verified alternatives list that the pharmacist presents to the prescriber for approval, compressing a multi-hour workflow into minutes while leaving authority and liability exactly where they legally must remain.
⏰Why Now
Three converging events opened this window specifically in 2023–2025: the FDA upgraded its shortage database API to enable programmatic real-time access; the GLP-1 shortage made drug shortage management a priority for every health system CFO watching drug spend explode; and updated CMS reimbursement rules now allow pharmacists to bill for clinical consultations around shortage management, creating a new revenue line that directly justifies software investment from health system procurement.
The Surprising Insight
US drug shortages hit a 22-year high in 2023, yet the standard clinical workflow for finding alternatives is still a pharmacist manually calling competitor pharmacies and checking FDA.gov lists — a process that has not changed since the 1990s and now consumes hours that patients no longer have.
Kill Reason
Critical weakness: Regulatory risk
What do you think?
Related ideas you can explore free:
killed: Urine and stool analysis for glucose monitoring, cancer screening, and kidney function markers falls squarely into FDA Class II/III medical device territory, requiring clinical trials and 510(k) or PMA approval costing millions and taking years. The entire high-value use case — the reason anyone would buy this — is legally undeliverable without regulatory clearance the product cannot realistically obtain.
killed: Amazon Alexa Together and comparable remote family monitoring services already serve this exact use case at consumer price points with embedded distribution advantages that a standalone WiFi sensing module cannot overcome. Daily routine monitoring and anomaly alerting for elderly relatives is a feature, not a product, in the current competitive landscape.
killed: Personal injury narrative reports require specific medico-legal structure, but the format is well-documented in public templates and any chiropractor with a configured AI prompt can replicate the output — there is no proprietary data asset, workflow lock-in, or network effect to justify a subscription fee against a commodity capability.