Hello, I'm Aditya
Aditya Gambhir
Software engineer with 3+ years building production-grade microservices and event-driven backend systems, bridging generative AI with distributed architecture to deliver secure, reliable software.
/ about
Engineer who ships across the whole stack
Software Engineer with 3+ years building production-grade microservices and event-driven backend systems processing enterprise-scale transaction volumes. Reduced release cycles from weeks to days through disciplined CI/CD, rigorous test-driven engineering, and 85% code coverage standards. Committed to bridging generative AI capabilities with distributed architecture to deliver secure, reliable, and high-performance systems.
Software Engineer (Backend & Distributed Systems)
- ▹Engineered a Kafka-based event streaming pipeline syncing member eligibility and claims data in real time, processing over 100,000 events daily with zero data loss
- ▹Migrated a multi-tenant clinical platform to one physical database per clinic office, validating isolation with a harness that disables tenant filtering
- ▹Cut payment processing time by 40% for a system handling over 500,000 daily transactions by rewriting joins and adding indexes
- ▹Built deployment pipelines with Docker, Jenkins, and Kubernetes, cutting release time from 2 weeks to 3 days
AI / LLM Engineer
- ▹Built an AI-powered member-support chatbot with Spring Boot, Azure AI, and LLM-based knowledge search, cutting agent response time by 22% and documentation search time by 35%
- ▹Ran a five-method bake-off for medical-record segmentation against hand-labeled ground truth; sliding-window won on Document-F1, recall, and cost
- ▹Architected a multi-agent literature-review system with Microsoft AutoGen and an LLM provider layer that falls back across Together AI, Gemini, Perplexity, and OpenAI
/ skills
The toolbox
Languages
Frameworks
Data / AI
Cloud & DevOps
Databases
Tooling
/ projects
Selected work
Full-stack geospatial navigation platform with real-time interactive mapping, multi-modal route planning, and live transit tracking for Olympic venue navigation, backed by a Spark/Hadoop pipeline processing 1M+ records a minute. Built with a 4-person team.
Multi-agent literature-review system: AutoGen search and summarizer agents automate arXiv retrieval, analysis, and structured review generation, behind an LLM provider layer with fallback across Together AI, Gemini, Perplexity, and OpenAI.
Admin Dashboard
Enterprise MERN dashboard with JWT auth, role-based access and real-time KPI charts.
/ experience
The journey so far
Software Engineer
Healthcare Support, LLC
- ▹Rebuilt a 7-year-old medical scheduling platform on .NET 10, ABP Commercial, and Angular 20, deploying to a self-hosted Ubuntu server, producing 846 commits, 390 pull requests, and 17 ADRs total.
- ▹Migrated the platform from shared-database row filtering to one physical database per clinic office, validating isolation with a test harness that disables tenant filtering and confirms cross-office records stay invisible.
- ▹Built two transactional outboxes using SHA-256 idempotency keys and atomic database-lease claiming, plus reconciliation jobs with per-office error isolation, ensuring notifications are never lost or duplicated under concurrent worker processes.
- ▹Ran a five-method bake-off for medical-record segmentation against hand-labeled ground truth; sliding-window won on Document-F1, recall, and cost, rejecting four alternative methods including text-based OCR consistently using fewer API calls.
Software Engineer
Centene
- ▹Built an AI-powered chatbot using Spring Boot, Azure AI, and LLM-based knowledge search to automate member support inquiries, cutting agent response time by 22% and documentation search time by 35%.
- ▹Developed REST APIs with OAuth2 authentication for internal and external member-facing systems, maintaining strict HIPAA compliance and 99.9% uptime across critical production healthcare environments and backend services daily.
- ▹Engineered a Kafka-based event streaming pipeline syncing member eligibility and claims data in real time across multiple internal healthcare systems, processing over 100,000 events daily with zero data loss.
- ▹Wrote unit and integration tests using JUnit and Mockito for backend services, maintaining approximately 85% code coverage and catching defects before they reached the production release pipeline each week.
- ▹Collaborated in Agile Scrum teams with Product, QA, and DevOps engineers, delivering new features in two-week sprints while conducting peer code reviews to maintain team engineering standards throughout each release.
Software Engineer
CRED
- ▹Developed 15+ REST APIs using Java and Spring Boot, writing automated JUnit tests to catch bugs early and reduce production issues before code changes reached the live payment processing system.
- ▹Optimized slow-running SQL queries by rewriting joins and adding database indexes across services, cutting payment processing time by 40% for a system handling over 500,000 financial transactions each day.
- ▹Configured AWS auto-scaling and access controls for backend services supporting real-time transaction processing, improving system uptime to 99.5% while reducing overall monthly cloud hosting infrastructure costs by 25%.
- ▹Refactored key ReactJS components by introducing Redux state management and removing unnecessary repeated API calls, cutting page load time for 50,000 monthly active users on the app.
- ▹Built 2 deployable microservices with automated unit tests in Mockito and created automated deployment pipelines using Docker, Jenkins, and Kubernetes, cutting release time from 2 weeks to 3 days.
Education
M.S. Computational Data Science
3.67/4University of California, Riverside · Mar 2025
B.Tech Computer Engineering
9.62/10Vishwakarma Institute of Information Technology · May 2023
Certifications
+ 4more — Coursera, Infosys & Udemy
A Comprehensive Survey of Multiple Object Tracking Techniques
Surveys object detection, tracking algorithms, and semantic segmentation for traffic-law enforcement — covering occlusion, splits, and deep-learning methods.
IJISAE · Read paper/ contact