IMPORTANT LOGIN FOR TRY-OUT
username: demo@tryvera.com password: 12345 (please don't change the password) https://github.com/Deep-Context-AI/vera-platform
Inspiration
Every month hospitals lose millions because organizations can’t onboard practitioners until their credentials clear. Manual data entry and pulling from disjointed sources lead to 15+ day timeframes for a single file (and a nightmare to organize). Hospitals lose ≈ US $9 k in revenue per provider for every day their credentials are stuck in limbo. Multiply that by the 90-120 days a typical file sits in queue and you get a multi-billion-dollar bottleneck. Vera validates and flags provider applications for human review in compliance with industry standards. Built to work WITH humans, NOT replace.
"After half a decade living that headache myself, I pictured a world where onboarding a clinician was as quick as spinning up a cloud server. That vision became Vera." - Annie, Credentialing Manager
How we built it
Vera uses the OpenAI Agents SDK and smart element selection to validate and automate data-entry through our mocked API hosted on modal. Bolt got us a great foundation with NextJS for our platform and 100% built our /marketing page, sign-in page (and Supabase connection), and platform dashboard. We then polished the table and verification page to work with GPT-4.1 whose decision trees are modeled after real-world credentialing procedures.
We generated 15K unique healthcare professionals modeled after real, public sources like the DCA California Medical Board API and NPI. Google's Gemini 2.5-flash helped us generate thousands of scenarios for these fake providers like full malpractice cases. (It was only ~$5 with +5M tokens! 😅).
Built With
- bolt.new
- cursor
- gemini
- modal
- nextjs
- official-dca-license-api
- official-npi-api
- openai
- openai-agents-sdk-js
- python
- typescript
- vercel
Log in or sign up for Devpost to join the conversation.