SeaSure: AI Agents for Boat Maintenance (Client Project) by Tran Tien VanSeaSure: AI Agents for Boat Maintenance (Client Project) by Tran Tien Van

SeaSure: AI Agents for Boat Maintenance (Client Project)

Tran Tien Van

Tran Tien Van

Client project. Built by Van Data Team for SeaSure, a subscription platform for boat owners that pairs AI troubleshooting with a document vault and a vetted service-provider directory. Live product: seasure.ai

The challenge

Boat owners let scheduled maintenance slip because troubleshooting knowledge and service history lived in different places.

What we built

A FastAPI backend of eight domain modules with a specialised LangGraph agent per boat subsystem, routing across Claude 3 and GPT-4
A React 19 + TypeScript frontend of roughly 25 pages on TanStack Query
PostgreSQL with pgvector, 25+ indexes, tuned connection pooling and Alembic migrations
Stripe with three tiers and usage quotas; JWT with Google OAuth and refresh-token rotation
WebSocket chat and live video on Google Cloud Run, with SendGrid and Twilio alerting

Results

A specialised LangGraph agent per boat subsystem
Three subscription tiers with usage quotas and rotating OAuth sessions
Real-time chat and live video over a pgvector-backed service directory

Key decision

One agent per subsystem rather than one agent for the whole boat: per-subsystem routing keeps each agent narrow enough to stay accurate.
Stack: FastAPI, LangGraph, Claude 3, GPT-4, React 19, TypeScript, PostgreSQL, pgvector, Stripe, Google Cloud Run, Twilio, SendGrid.
Full case study: vandatateam.com/portfolio
Like this project

Posted Nov 5, 2026

Client project for SeaSure: FastAPI + LangGraph agents per boat subsystem, React 19 app, pgvector service directory, Stripe tiers, live chat and video.