Deepika Chintamreddy
Hyderabad, India
deepikachintamreddy447@gmail.com
Full-Stack Developer Intern
ISRO, Satish Dhawan Space Centre (SHAR)
- Built the backend of a Network Switch Monitoring Dashboard (JSP, MySQL) with REST APIs for real-time analysis of switch status and performance across a large distributed network.
- Documented the architecture and API specifications.
Backend Engineering Intern
Neuroity
- Built a dataset search aggregator with Django REST Framework across 10 dataset platforms, using asynchronous fan-out, Redis-backed caching, Celery background jobs and JWT authentication.
- Debugged a data-provider integration issue and cleaned up the codebase to deliver a production-ready handoff package via Git.
GPU-accelerated bus reliability analytics built for the APAC GenAI Academy (NVIDIA track). Processes 150M+ bus records with a 39.39x speedup using RAPIDS cudf.pandas, with a Gemini-powered copilot on top.
Multi-Modal Evidence Review System
Gemini 2.5 Flash vision-language pipeline that classifies claim status from image and video evidence, reaching 94.9% claim-status accuracy.
MCP server powered by Claude Haiku that automates healthcare prior-authorization over FHIR/SHARP clinical data, turning unstructured requests into structured, evidence-linked decisions. Includes the A2A protocol and Docker; submitted to Devpost Agents Assemble Challenge.
MCP server for family care coordination, built for the Amazon Developer Hackathon (Alexa+ track). Tools to log doses, check what is due, prepare doctor-visit briefs, request refills and erase stored data. Refill orders need explicit confirmation, with token authentication and role checks.
Multi-Domain Support Triage Agent
Triage agent built for HackerRank Orchestrate that routes support tickets across three domains. Retrieves evidence with hybrid BM25 and dense search plus cross-encoder reranking, and escalates low-confidence cases instead of guessing.
Spatio-temporal traffic risk model built for Gridlock Hackathon 2.0 on Bengaluru Traffic Police incident data. Captures 68.7% of congestion, cuts MAE by 33%, and identifies about 21% of congestion-minutes as preventable.
M.Tech (Artificial Intelligence & Data Science)
Alliance University
Attendee · Oct 24, 2026