Multi-Tenant CRM & Activity Graph
An enterprise customer relationship system built around a dynamic resource-allocation graph for matching clients with specialized tutors/advisors.
Year
2024
Services
Architecture Scope
- —Dynamic relationship mapping
- —Heuristic matching algorithm
- —Timezone scheduling matrix
- —Automated invoice generation
- —Unified progress timelines
20K+
Active Client Profiles
Globally distributed
95%
Matching Efficiency
Automated schedule pairing
12+
Countries Supported
Real-time timezone sync
100%
Automated Invoicing
Zero manual intervention
The Challenge
Managing student intake, skill level evaluations, and certified tutor pairings across different global timezones often leads to double-bookings and billing errors. COORDINATING scheduling sheets, tutor certifications, and multi-currency billing manually represents a major bottleneck for educational and advisory platforms.
The Solution
We designed a multi-tenant CRM that models scheduling as a bipartite graph of availability and credentials. The platform automates intake, parses tutor skill profiles, maps timezone availabilities, and reconciles invoicing automatically for users in multiple countries.
System Architecture
Built on NestJS and PostgreSQL, the platform leverages Prisma ORM. A customized scheduling solver computes intersection sets of tutor availability intervals in real-time. A webhook state machine handles billing events with Stripe, facilitating automated class auditing and payroll accounting.
Technology Stack
- React
- NestJS
- PostgreSQL
- Prisma ORM
- AWS S3
- Twilio
- Stripe