Antariksh Saathi
A retrieval-augmented conversational system for satellite data, shaped around trustworthy answers and reduced hallucination.
- RAG
- Python
- Vector DB
AI systems, backend products, and business automations—designed to survive outside the demo.
00 / SIGNAL
01 / SELECTED SYSTEMS
A retrieval-augmented conversational system for satellite data, shaped around trustworthy answers and reduced hallucination.
A scalable analysis pipeline and API layer for a deep-learning framework that detects land-use changes from satellite imagery.
Privacy-conscious backend infrastructure for an AI-assisted wellness and mental-health support platform.
02 / HOW I BUILD
RAG pipelines, conversational agents, deep-learning workflows, evaluation, and the guardrails needed to make outputs dependable.
Python / PyTorch / TensorFlow / Vector DBsAPIs, data flows, database-backed services, and integration layers built around the product—not around a technology showcase.
Node.js / MongoDB / REST APIsPractical systems that connect forms, research, data, and follow-up so teams spend less time moving information by hand.
Workflows / Internal tools / Web apps03 / CONTEXT
I am a Computer Science engineering student at Chandigarh University, building at the intersection of AI, backend engineering, and real operational problems.
The recurring theme is translation: turning difficult models into understandable outputs, user feedback into safer systems, and unclear business bottlenecks into small tools that can ship.
Outside project work, I represent 100+ students as a class representative. That has made communication, ownership, and honest expectation-setting part of how I engineer—not a line added afterward.
04 / COMMERCIAL WORK
RichKeed is where I build sales websites, workflow automations, and focused internal tools for service businesses and small agencies.
Visit RichKeed ↗For a defined build, a technical collaboration, or a conversation about where your system is stuck.