Aide

Inspiration

After leaving the hospital, patients—especially elderly individuals—are often given complex discharge instructions and expected to manage everything on their own. This leads to missed medications, delayed follow-ups, and preventable complications. We were inspired to solve the gap between medical instructions and real-life follow-through, particularly for patients who struggle with technology.

What it does

Aide is a voice-first AI care companion that turns discharge instructions into a simple, structured care plan. It actively checks in with patients through voice, tracks whether they complete tasks, and adapts when something goes wrong.

Converts medical instructions into clear daily tasks Uses voice to check in with patients (elderly-friendly) Detects missed care and provides safe next steps Identifies risk and generates a doctor-ready summary

Aide ensures that care doesn’t stop when the patient leaves the hospital.

How we built it

We built Aide using v0, Next.js, and Tailwind CSS to quickly create a clean, accessible healthcare interface.

We used an agent SDK to structure the care coordination flow, and Grok for reasoning over patient inputs, missed tasks, and escalation scenarios.

Key components: v0 for rapid UI generation Next.js for the web app Tailwind CSS for styling Agent SDK for care plan workflow logic Grok for reasoning and response generation Voice interaction for elderly-friendly check-ins Doctor summary generator for escalation cases

We focused on building a real-time, interactive demo that shows how Aide can guide patients after discharge, check in through voice, adapt when care is missed, and prepare a doctor-ready summary when risk is detected. We used Vercel to quick deploy it.

What's next for Aide

Build enterprise healthcare integrations so doctors and care teams can directly update a patient’s post-op care plan and discharge instructions Connect with pharmaceutical services to automate refills, medication delivery, and pharmacy coordination Add real phone calls with a human-in-the-loop option for higher-risk patients Integrate real LLM-based document parsing for any discharge note Enable caregiver notifications and alerts Expand into chronic care management and remote patient monitoring

Built With

• css
• grok
• next.js
• sdk
• tailwind
• v0