Product Thesis
The useful layer is not another dashboard. It is the work surface that turns signal into action.
Small real-estate teams already juggle CRMs, spreadsheets, calendars, inboxes, and social tools. Vyndir’s product argument is that the missing layer is operational clarity: what changed, what matters, who owns it, and what should happen next.
The demo focuses on a seeded brokerage team so reviewers can inspect a complete workflow model instead of a shallow landing page or a generic analytics shell.
Built Product
The app has enough working surface area to evaluate real execution.
Configurable dashboard with vertical-aware KPI packs
Relationship workspace for clients, leads, listings, deals, and appointments
CSV import preview, mapping, validation, processing, and metric refresh
Grounded AI brief panels attached to operating records
Worker architecture for imports, metrics, email, notifications, and marketing schedules
Architecture
Full-stack shape behind the demo.
The architecture is intentionally modern TypeScript: a web app, API, relational schema, worker layer, realtime foundation, and shared contracts that can be inspected as a coherent system.
Web
Next.js App Router with typed server data access and shared UI primitives.
API
NestJS REST API with cookie auth, role-aware workspace routes, and Swagger docs.
Data
PostgreSQL schema through Prisma with real-estate entities and import history.
Workers
BullMQ and Redis workers for background processing and operational fanout.
AI Direction
AI is framed as grounded decision support, not generic chat.
The portfolio demo keeps deterministic AI briefs so the public version does not need third-party model credentials. The product pattern is still visible: AI appears inside records and dashboards, cites source context, and drafts next actions for human review.
That design choice shows judgment around trust, auditability, data readiness, and the difference between a useful copilot and a thin chatbot.
Five-minute reviewer path.
Dashboard: see the daily operating brief, AI recommendations, and chosen KPIs.
Leads/Listings/Deals: inspect cross-linked real-estate records and AI briefs.
Imports: review CSV preview, field mapping, validation, and job history.
Marketing Assets: see listing-attached content planning as a supporting workflow.
Settings/Operations: review runtime readiness and portfolio-demo constraints.