OBJX: Multi-Agent AI Platform with Memory (Client Project)
Tran Tien Van
Client project. Built by Van Data Team for OBJX Intelligence, the AI platform of OBJX Design, an architecture and design practice in San Diego. Live product: meetobjx.com
The challenge
The platform needed agents that remembered prior work across sessions while staying isolated per tenant.
What we built
A five-agent architecture on LangGraph and LangChain over Claude Sonnet and GPT-4, with MEM0 long-term memory
FastAPI on PostgreSQL 15, Redis, async SQLAlchemy 2.0, Alembic and Celery
A React 18 + TypeScript frontend on Vite with shadcn/ui and TanStack Query
A five-tier permission system enforcing multi-tenant isolation
Six enterprise integrations over OAuth2 and webhooks: Google Workspace, Slack, Teams, Salesforce, Monday.com and Asana
Results
94.7% platform health score in production
Five specialised agents with MEM0 long-term memory
Five-tier permission model with multi-tenant isolation
Key decision
Isolation is enforced in the data layer rather than in agent instructions, because an instruction is not a boundary.
Stack: LangGraph, LangChain, MEM0, Claude Sonnet, GPT-4, FastAPI, PostgreSQL 15, Redis, Celery, React 18, TypeScript, DigitalOcean.
Full case study: vandatateam.com/portfolio
Like this project
Posted Nov 5, 2026
Client project for OBJX Intelligence: five LangGraph agents with MEM0 memory, multi-tenant FastAPI + React platform, 94.7% health score in production.
