Inspiration
A drug passing clinical trials is only step one. Most countries then ask: "Is this treatment worth the money?" Regulators require formal economic evidence before covering a new therapy. Without it, an approved drug never reaches patients. Building that evidence costs $50Kâ$200K+, takes months, and requires rare expertise. In most low- and middle-income countries, no such analysis exists at all. I built GENESIS to make this process fast, affordable, and accessible to anyone.
What it does
GENESIS automates cost-effectiveness analysis. A user selects a disease, country, treatment, and comparator â and within minutes receives a complete analysis with a publication-ready report and full citations. It covers 5 chronic diseases across 190+ countries, using published clinical data adapted to each country's demographics, costs, and health system.
How we built it
-replicating peer reviewed and published health economic models within GENESIS, and making it adaptable to any country so that its insights are locally relevant and credible.
Challenges we ran into
-Adapting to 190+ countries. Each analysis uses local life tables, disease prevalence, purchasing-power-adjusted costs, and country-specific spending thresholds. Edge cases are everywhere. -Speed without shortcuts. Every parameter traces to a published source, every model is validated against real-world targets, and uncertainty is quantified across 1,000 probabilistic iterations. Making this run in minutes instead of months was the core engineering challenge. -Making results understandable. The hardest problem is communication, not computation. Getting Nova's prompts right â accurate, contextual, and readable by a non-specialist â required significant iteration.
Accomplishments that we're proud of
Generating the same findings as recently published and independent studies in leading medical journals, in minutes, not months or years.
What we learned
Multimodal AI is most powerful not when generating new analysis, but when translating validated results into language that decision-makers can act on.
Log in or sign up for Devpost to join the conversation.