Multi-tenant fleet analytics with RAG-powered intelligence.
Multi-tenant
Architecture
RAG
AI Retrieval
RBAC
Access Control
T3 / tRPC
Stack
Overview
At MineSync.ai, an analytics startup for industrial mining fleets, I worked across the full stack to build enterprise-grade, multi-tenant software. The product ingests operational data from mining fleets and surfaces it through analytics dashboards and AI-assisted workflows.
I developed the core Lifecycle Management workflows using TypeScript, Next.js, Prisma, and tRPC, with tenant-isolation middleware and scalable data filtering to keep each customer's data cleanly partitioned. On the AI side, I implemented RAG-powered LLM information retrieval for fleet analytics, layered behind granular role-based access control (RBAC) so different roles saw only what they were permitted to.
I also focused on performance — optimizing API queries and React dashboard visualizations to eliminate redundant network calls and reduce dashboard response latency, improving the experience for operators working with large, live datasets.
What I did
- —Developed core Lifecycle Management workflows for mining fleet analytics using TypeScript, Next.js, Prisma, and tRPC with tenant-isolation middleware and scalable data filtering
- —Implemented RAG-powered LLM information retrieval for fleet analytics with granular RBAC permission layers across multi-tenant deployments
- —Optimized API queries and React dashboard visualizations, reducing redundant network calls and improving dashboard response latency
Tech stack
Languages
Full-Stack
AI & Security
Connect with Dhruv Hegde
See more of Dhruv Hegde's work and background on LinkedIn, GitHub, and ResearchGate.