Inspiration
When my Mom had a fall and was admitted to the ICU, she was prescribed a heavy dosage of medication that, unfortunately, worsened her condition.
As patients or caregivers, we often trust prescriptions without questioning them because verifying medical advice can be difficult and time-consuming. Searching online or consulting multiple doctors is not always practical in urgent situations.
This experience made us realize that patients and families need a quick, reliable way to validate prescriptions and understand potential medication risks. That moment inspired us to build Aarogya Health Assistant—a system designed to help people quickly verify prescriptions, detect possible medication errors, and make more informed healthcare decisions.
What it does
Voice in Your Language – Users can speak in Hindi, English, or any other language. Aarogya understands them and responds, eliminating the need for typing or reading.
Prescription Safety Check – Users upload their prescriptions, and Aarogya checks for dangerous drug combinations, missing dosages, incomplete details, and high-risk medicines before taking any medication.
Ayurvedic + Home Remedy Alternatives – Suggests trusted Indian home remedies and Ayurvedic alternatives for conditions, offering natural options for those who prefer them over conventional medicine.
Global Relevance – Aarogya is accessible to users in India and abroad, combining modern medicine with ancient wisdom, reflecting the global interest in natural healing.
Alexa Reminder – Never Miss a Medicine – Automatically sets medicine reminders on Alexa devices after analyzing the prescription, helping caretakers manage elderly patients or those with multiple medications.
Essentially, our system combines voice interaction, prescription safety, natural alternatives, global accessibility, and smart reminders in one platform.
How we built it
We built Aarogya Health Assistant by starting with the Nova Sonic sample code provided by Amazon Web Services and customizing it to address prescription safety challenges. First, we modified the user interface to allow users to upload prescription images for analysis. We then enhanced the AI prompts to provide better healthcare context specific to Aarogya, enabling more accurate responses.
Next, we expanded the system’s knowledge base so the AI agent could analyze prescriptions, detect potential issues, and distinguish between safe and potentially problematic prescriptions. We also incorporated additional medical knowledge to allow the system to suggest alternative medicines where appropriate.
Finally, we integrated the solution with Amazon Echo devices so that once a prescription is verified, the system can automatically set medication reminders through Alexa, helping patients—especially elderly users—remember to take their medicines on time.
Challenges we ran into
Decoding Sample Code: Figuring out how the starter code worked and how to adapt it.
Prescription Validation: Identifying what makes a prescription “good” or “bad.”
Knowledge Base Integration: Plugging our data into the agent-core system.
Alexa Integration: Getting our solution to work seamlessly with an Alexa Echo.
Accomplishments that we're proud of
Learned Nova Sonic inside-out
Hooked up the drugs & alternative medicine knowledge base
Made it Alexa-ready
What we learned
How to navigate and adapt the sample code quickly.
How to evaluate “good” vs. “bad” prescriptions.
Techniques for integrating a knowledge base into the agent core.
Challenges and solutions for connecting our system with the Alexa service.
Coordinate with team members and advisors across the globe.
What's next for Aarogya Health Assistant
Our next goal is to secure startup support and AWS cloud credits from Amazon Web Services to scale Aarogya Health Assistant, expand its AI capabilities, and deploy it for real-world healthcare use.
With this support, we aim to improve the AI models, expand the medical knowledge base, and integrate more healthcare services such as multilingual support, deeper prescription analysis, and better doctor–pharmacy communication. Ultimately, our vision is to make Aarogya a reliable digital health assistant that improves medication safety and accessibility for patients across India.
Built With
- agent
- agent-core
- apig
- bedrock
- dynamodb
- lambda
- python
- typescript
Log in or sign up for Devpost to join the conversation.