Chikit.AI
An end-to-end AI-powered healthcare platform built from 0 to 1 — covering intelligent patient assessment flows, multi-enterprise RBAC dashboards, real-time discharge automation, and a progressive web app for frontline clinical workflows.

The Impact
Full Platform Ownership
Defined information architecture, user flows, and component system from scratch.
Enterprise Portals
Separate dashboards for Hospitals, Clinics, and Super Admins with RBAC controls.
First-of-Kind Flows
Designed novel interfaces for LLM-powered discharge summaries and patient file analysis.
The Challenge.
Designing for life-critical workflows.
Healthcare systems are notoriously fragmented. Clinicians must navigate multiple tools — patient records, assessment forms, discharge documents — resulting in wasted time and critical errors.
Chikit.AI needed a single, intelligent platform that could consolidate these workflows while introducing AI assistance without overwhelming clinical staff who may not be tech-savvy.
Key Design Problems
Platform Modules
AI Assessment Flow
Conversational symptom capture with AI-generated preliminary reports. Designed for both in-clinic tablets and patient self-serve kiosks.
Patient File Analysis
Upload any document — lab report, scan, prescription — and the system extracts and structures the data automatically, surfacing key insights to the doctor.
Discharge Summary Automation
AI generates discharge summaries based on visit data and doctor notes. Doctors review, amend, and approve — not write from scratch.
Enterprise RBAC Dashboard
Multi-tenant management for hospital chains. Super-admins manage clinics, access levels, billing, and analytics across all locations.
Assessment & Patient Flow
End-to-end patient journey from intake to discharge.
Drop your Figma export here
Replace with actual screenshot
Clinical Dashboard
Doctor-facing dashboards with AI-assisted diagnosis tools.
Drop your Figma export here
Replace with actual screenshot
Enterprise Management
RBAC admin panel for multi-clinic management.
Drop your Figma export here
Replace with actual screenshot