Ronit Khanna

Projects

HomeaZZon – AI-Powered Property Management (Capstone)

An enterprise-grade property management platform developed under academic–industry constraints, focused on backend system design, document automation, and predictive maintenance workflows.

  • Designed and implemented a modular .NET 8 backend service integrating Azure Form Recognizer for document ingestion
  • Built a structured data extraction pipeline to process invoices and warranty documents from scanned inputs
  • Defined DTO-based service contracts to enforce clear boundaries between controllers, services, and data layers
  • Designed a predictive maintenance workflow with controller stubs and extensible data flow for future ML integration
  • Integrated Azure AI services with Azure SQL Server while maintaining separation between business logic and AI providers
  • Built as part of an academic–industry capstone with enterprise architecture and scalability constraints

Tech: .NET 8, C#, Azure AI Vision, Azure Form Recognizer, Azure SQL, Angular

View on GitHub →

Async Job Processing System

A distributed asynchronous job processing system designed to handle long-running tasks with retries, idempotency, and fault tolerance.

  • Designed a Redis-backed job queue enabling asynchronous task submission and background execution
  • Implemented distributed locking to guarantee at-most-once job execution across multiple worker instances
  • Built idempotency handling to prevent duplicate job submissions under client or network retries
  • Implemented automatic retry logic with exponential backoff and failure recovery
  • Designed horizontally scalable worker processes for parallel job execution
  • Integrated OpenAI-powered tasks while keeping AI logic decoupled from core job orchestration

Tech: Python, FastAPI, Redis, OpenAI API

View on GitHub →

Distributed URL Shortener

A production-ready URL shortening service designed for high throughput, low latency, and safe API usage.

  • Designed RESTful APIs for URL creation and redirection with sub-50ms average latency
  • Implemented Redis-based caching achieving 99%+ cache hit rate under load
  • Built token bucket rate limiting to prevent abuse and ensure fair API usage
  • Designed Base62 encoding strategy for compact, collision-resistant short URLs
  • Containerized backend services using Docker for consistent deployment and scaling
  • Designed to simulate real-world traffic patterns and backend scalability constraints

Tech: Java, Spring Boot, Redis, PostgreSQL, Docker

View on GitHub →

Car Rental Application

A backend-driven car rental system designed to model real-world booking flows, authentication, and administrative control paths.

  • Designed RESTful APIs to manage bookings, users, vehicles, and availability with clear domain boundaries
  • Implemented role-based access control to separate customer-facing operations from admin workflows
  • Applied backend-first design principles to ensure data consistency across booking and inventory operations
  • Designed to simulate a production-grade booking system with administrative and operational constraints

Tech: Node.js, Express, MongoDB, React

View on GitHub →

AI Study Assistant

A backend-focused AI application designed to deliver context-aware study responses using prompt-driven APIs.

  • Built a FastAPI backend to handle request orchestration and AI-driven response generation
  • Designed prompt handling and API integration to keep model interactions isolated from core business logic
  • Implemented MongoDB-based persistence for session-level context and conversation state
  • Focused on scalable AI API integration patterns rather than model-specific logic

Tech: Python, FastAPI, MongoDB, OpenAI API

View on GitHub →

EV Charging Finder (iOS)

A native iOS application built to explore real-time location-based data access and API-driven mobile workflows.

  • Implemented MVVM architecture using SwiftUI to separate view logic from data and network layers
  • Integrated OpenChargeMap API to fetch and display real-time EV charging station data
  • Designed a map-first user experience with search and favorites for repeated access
  • Built to practice production-relevant iOS patterns including API integration and state management

Tech: Swift, SwiftUI, OpenChargeMap API, iOS

View on GitHub →