AI Platform

Summary We run an AI-powered SaaS platform (Next.js/NestJS, PostgreSQL) that uses LLM calls (RAG, structured outputs) across multiple client tenants. We currently have no visibility into per-tenant AI usage or cost — we need a dashboard that tracks and displays: Token usage per tenant/client, per feature (e.g. quote generation, document parsing) Estimated cost per request and aggregated per day/week/month Basic breakdown by AI operation type (e.g. RAG query vs structured extraction) A simple admin-facing UI to view this data (charts + table view) Scope: Backend: middleware/logging layer to capture token usage + cost per LLM call, stored in PostgreSQL Frontend: an admin dashboard view (charts, filters by tenant/date range) Should integrate cleanly with our existing Next.js/NestJS stack Ideal candidate: Experience with LLM integration (OpenAI/LangChain), Next.js, NestJS, PostgreSQL, and building internal analytics/admin tooling.