Dawrak
Dawrak is a digital clinic receptionist Agent that answers approved FAQs, collects appointment requests, escalates to humans, and supports follow-up through dashboard operations. It is productized with bilingual UX, n8n production workflows, and integrated as a first-class Agent inside Assistants Pro AI.
Core Tech Stack
- Next.js
- React
- TypeScript
- n8n
- OpenAI
Performance
- Workflow source centralized for faster iteration and rollback
- Synthetic parity tests before live channel expansion
- Feature-flagged live chat to avoid accidental production exposure
- Reusable Agent packaging instead of rebuilding clinic bots per client
Security
- No unsupported medical claims in agent responses
- Lead form fails closed when webhook destination is unconfigured
- Secrets and channel credentials stay out of public artifacts
- Compat redirects and live flags remain default-off unless authorized
- Tenant/Agent isolation contracts on the platform layer
Key Features
Clinic reception positioning: FAQ + appointment request + human handoff
Safety boundaries against diagnosis/prescription claims
Beta access chat flows for controlled demos
Landing conversion page with lead webhook contract
n8n clinic workflow family transferred to platform agents/dawrak
Synthetic parity module on Assistants Pro AI (FAQ/appointment/handoff/locale)
Tenant Agent settings schema for clinic configuration
Marketing package with banners, scripts, and launch materials
MongoDB Atlas multi-tenant planning model (database-per-tenant)
Live chat wiring path through Assistants Pro AI platform flags
Telegram/operational logging paths for staff visibility
Arabic-first product branding with English platform shell support
Project Architecture
frontend/ # Clinic product web surface agent/workflows/clinic-source/ # Legacy reference workflows marketing/ # Launch creatives and campaign pack Assistants Pro: agents/dawrak/ # Authoritative workflow source Assistants Pro: platform-agent-dawrak package Assistants Pro: /agents/dawrak marketing + demo surfaces
Engineering Strategies
- Sell a business outcome (digital receptionist), not a generic medical chatbot
- Keep Sheets operational until a verified Atlas migration window
- Transfer workflow source authority into the platform with backup/rollback
- Use Assistants Pro AI for packaging, demos, and tenant workspaces
- Enforce honesty chips and safety guardrails in chat UX
Visual Interface
