Projects
A collection of noteworthy professional, freelance, academic, and personal projects I've worked on.
-
AI-Powered Product Recommendation System
Leading the development of an AI-powered product recommendation system designed to serve tens of thousands of field agents. The system leverages Retrieval-Augmented Generation (RAG) and agentic engineering patterns to deliver intelligent, context-aware product recommendations at scale. Built a React frontend for agent-facing workflows and integrated Databricks for large-scale data processing and MLflow for experiment tracking and model lifecycle management.
- Python
- RAG
- Databricks
- WebSockets
- MLflow
- Agentic Engineering
- React
- AI/ML
- Distributed Systems
-
Integrated Pest Detection and Alerting System
Integrated a pest detection and alerting capability into an internal device management system. Enabled real-time alerts and exposed device data through RESTful API endpoints for enhanced monitoring. Also, developed a Blazor-based dashboard for visualizing pest detection data and device status. Added unit and integration tests to ensure system reliability and performance.
- C#
- .NET Core
- Azure
- REST API
- Microservices
- Blazor
- Distributed Systems
-
Executive Business Review Web Application
Co-led the development, delivery, and feature enhancements of an EBR web application used by 300+ global account managers. The application standardized reporting across the organization and significantly improved customer engagement. Built with modern web technologies and integrated with backend analytics services for real-time data visualization.
- Angular
- TypeScript
- C#
- .NET
- Snowflake
- Azure SQL
- Redis
- CI/CD
-
Migrated Legacy Email Notification System to Event-Driven Architecture
Migrated legacy email notification processing from a monolithic system to an event-driven architecture. Implemented asynchronous processing using Azure Functions and Service Bus for improved scalability and reliability. Reduced notification processing time by 70% and enhanced system maintainability.
- C#
- .NET Core
- Azure Functions
- Service Bus
- Distributed Systems
-
Enabled Observability of Commercial Systems through APM Tools
Worked with architecture teams to onboard commercial systems to Application Performance Monitoring (APM) using AppDynamics and Application Insights for improved visibility and troubleshooting. Implemented monitoring solutions to track system performance and identify bottlenecks. Enhanced system reliability and reduced downtime through proactive monitoring.
- C#
- .NET Core
- AppDynamics
- Application Insights
- Distributed Systems
-
Remote Cache Server for JS Monorepo
Developed a remote cache server to optimize build times for a large Javascript monorepo. Implemented caching strategies to store and retrieve build artifacts, significantly reducing build times for developers by up to ~90%. Improved developer productivity and streamlined the CI/CD pipeline.
- Python
- FastAPI
- Docker
- Blob Storage
-
Microfrontend and UX Library Infrastructure Setup
Designed and implemented infrastructure for a microfrontend platform and reusable UX component library. Set up CI/CD pipelines to automate builds, testing, and deployments. Configured Terraform for infrastructure-as-code management of Azure resources. Implemented blob storage for artifact management and Front Door CDN for optimized global content delivery. This DevOps work enabled multiple teams to develop and deploy independently while maintaining shared UI components across applications.
- Terraform
- CI/CD
- Azure
- Azure Blob Storage
- CDN
- Microfrontends
-
AI Innovation Initiative - Developer Experience Experiments
Participated in an innovation initiative with data scientists to experiment with and prototype cutting-edge AI tools for improving developer experience. Prototyped four key AI applications before commercial alternatives became available: CLI Unit Test Generator for automating test creation, ScrumMasterAgent for managing sprint workflows, Software Engineering Insights App for analyzing code metrics and team productivity, and Code Review Agent for intelligent code review feedback. These early-stage experiments explored how AI could streamline engineering workflows and enhance developer productivity.
- Python
- AI/ML
- LLMs
-
AWS Infrastructure-as-Code for Multi-Million Commercial Product
Designed and implemented comprehensive infrastructure-as-code (IaC) for AWS resources supporting a multi-million dollar commercial product. Utilized Terraform to define, provision, and manage cloud resources reproducibly across development, staging, and production environments. Implemented best practices for security, scalability, and cost optimization. Automated infrastructure deployment and enabled reliable, consistent infrastructure management across teams.
- Terraform
- AWS
- CI/CD
-
Azure Data Factory Pipelines for Field Operations
Designed, modified, and maintained complex data pipelines in Azure Data Factory for field operation sales and service data processing. Handled high-volume data ingestion and transformation of hundreds of thousands of records daily. Implemented robust error handling, data validation, and monitoring to ensure reliable data flow. Optimized pipeline performance and maintained data quality standards across the enterprise data platform.
- Azure Data Factory
- ETL
- Azure SQL
-
Databricks Adoption for International Data Ingestion
Led the adoption and implementation of Databricks to significantly increase compute power for ingesting and processing international data sources. Architected a scalable solution that enabled faster data processing, improved performance, and reduced processing time. Coordinated cross-functional teams to migrate workloads and established best practices for leveraging Databricks' distributed computing capabilities.
- Python
- Databricks
- Apache Spark
- ETL
- Distributed Systems
-
Billing Invoice Processing Cron Job System
Co-built and implemented along with a principal software engineer a high-performance cron job system for processing thousands of daily billing invoices. Leveraged Spring Batch framework with multithreading to efficiently handle large-scale batch processing while maintaining data integrity. The system processed invoices in parallel, significantly reducing processing time and enabling timely billing operations. Implemented comprehensive error handling, retry logic, and monitoring for reliable invoice generation and delivery.
- Java
- Spring Batch
- Multithreading
- SQL
-
Business Partner Console Revamp
Contributed to the revamp of the business partner console, a critical portal used by thousands of business partners to manage partner relationships and operations. Worked on both frontend and backend development using React for the modern, responsive user interface and Spring Boot with RESTful APIs for robust backend services. Collaborated with product and design teams to deliver an improved experience that enhanced partner engagement and operational efficiency at scale.
- React
- TypeScript
- Spring Boot
- Java
- MySQL
-
DevSecOps Metrics Portal and Engineering Maturity Assessment Tool
Built an internal DevSecOps tool suite portal to showcase engineering team metrics and automate engineering maturity assessments. The platform aggregates data from multiple sources to provide comprehensive visibility into team performance, security posture, and development practices. Automated the assessment of engineering maturity across teams, identifying improvement areas and tracking progress. Enabled leadership with actionable insights and data-driven decision making for engineering excellence.
- React
- TypeScript
- Python
- FastAPI
- PostgreSQL
- DevSecOps
-
Playwright Test Suite Migration
Migrated legacy Selenium-based test suites to Playwright for a startup automation client. Significantly improved test reliability, runtime performance, and maintainability. Implemented parallel test execution and better error reporting.
- Playwright
- Python
- QA Automation
-
Government Links Auditing Tool
Developed a concurrent Python tool to rapidly audit third-party government links embedded within applications. The tool quickly detected broken links caused by government department closures during the DOGE (Department of Government Efficiency) initiative. Implemented concurrent processing to efficiently scan hundreds of links and identify failures, enabling swift remediation of broken references and ensuring application reliability.
- Python
- Concurrent Processing
- Web Scraping
-
Dynamic Report Generation Pipeline
Automated the generation of 80-100 Tableau reports using Selenium that originally required three resources working for four days to complete. Implemented intelligent report automation that reduced the entire process to just 3 hours, dramatically improving efficiency and freeing up resources for higher-value work.
- Python
- Selenium
- Tableau
- RPA
-
Selenium Test Framework with Page Object Model
Created a robust test automation framework using Selenium with Page Object Model (POM) design pattern to enable consistent and maintainable test creation across two core commercial applications. Implemented reusable page objects and utilities to reduce code duplication and improve test readability. The framework enabled the QA team to efficiently write and maintain automated tests for critical business workflows.
- Selenium
- Python
- Page Object Model
- Test Automation
-
Mobile Testing Framework with Appium
Led the development of a comprehensive mobile testing framework using Appium to automate testing of the mobile app's APK. Established best practices and patterns for mobile test automation, enabling the QA team to efficiently test mobile features across different devices and OS versions. The framework provided reliable mobile test coverage and accelerated mobile app validation.
- Appium
- Mobile Testing
- Python
- Test Automation
-
AxiomStream - Event Streaming Platform
Building AxiomStream, an event streaming platform written in Java and Spring Boot as a deep dive into distributed systems engineering. The project's ambitious goal is to explore core distributed systems concepts hands-on — including log replication, fault tolerance, leader election, and resilience patterns. Ongoing work in progress driven by a passion for understanding how systems like Kafka work under the hood.
- Java
- Spring Boot
- Distributed Systems
- Event Streaming
- Replication
-
Franz - AI Personal Assistant
Built Franz, a voice-driven AI personal assistant that integrates with home and system tools. Uses Whisper for speech recognition and Kokoro for text-to-speech, with Ollama powering a local LLM and LangChain orchestrating tool-use workflows. Franz can control smart home devices like the TV, retrieve system information, fetch live weather data, and more — all through natural voice interaction. Built with a React frontend for configuration and status display.
- Python
- React
- Whisper
- Kokoro
- Ollama
- LangChain
- AI/ML
-
NexusTerm - Rust Terminal Multiplexer
Building NexusTerm, a terminal multiplexer inspired by tmux with a modern desktop twist. Written in Rust for performance and safety, and packaged as a native desktop application using Tauri. The project explores terminal emulation, pane/session management, and systems programming while delivering a fast, cross-platform terminal experience.
- Rust
- Tauri
- Systems Programming
- Terminal Emulation
-
Personal Portfolio & Blog Website
Built this personal portfolio and blog using Astro and Tailwind CSS. Features include blog posts about learning and technology, a digital library of books I've read, and a curated list of projects. Designed for performance and SEO.
- Astro
- TypeScript
- Tailwind CSS
- SSG
-
DeltaView - Diff Checker Desktop App
Developed DeltaView, a desktop application for comparing and visualizing differences between files and text. Built with C++ and Qt framework, the application provides an intuitive interface for diff checking with syntax highlighting and side-by-side comparison views. A practical tool for developers and content creators who need to analyze file changes quickly and efficiently.
- C++
- Qt
- File Comparison
-
ProgLog - Distributed Append-Only Log Service
Built a distributed, append-only log service inspired by Apache Kafka. Implemented in Go, ProgLog provides a scalable messaging and event streaming platform for building distributed systems. The service handles high-throughput data ingestion, provides ordering guarantees, and enables multiple consumers to process events independently. A hands-on exploration of distributed systems concepts and protocols.
- Go
- Distributed Systems
- Message Queue
- Event Streaming
- Concurrency
-
SimLab - Simulation Job Service
Building SimLab, a Python service for submitting and managing simulation jobs through a REST API. The service accepts simulation parameters, executes jobs in the background, and provides endpoints for querying job status and retrieving results. Designed to handle asynchronous job processing and provide reliable result management. Work in progress.
- Python
- REST API
- Background Jobs
-
Hotel Management App
Built a comprehensive hotel management application for handling reservations, room management, and guest services. The application features a user-friendly interface and backend systems for managing bookings and customer data efficiently.
- Angular
- Java
- Spring Boot
- Multithreading
- SQL
-
Most Optimized Delivery Route
Developed an optimization algorithm to solve the vehicle routing problem and find the most efficient delivery routes. Used advanced algorithmic techniques to minimize travel time and distance while handling multiple constraints and delivery points.
- Python
- Graph Theory
- Nearest Neighbor
- Genetic Algorithms
-
Pollution Level Predictor
Created a machine learning model to predict pollution levels using Random Forest algorithm combined with gradient optimization techniques. The model analyzes environmental data and historical patterns to forecast air quality and pollution levels with high accuracy.
- Python
- Machine Learning
- Random Forest
- Gradient Optimization
-
Walmart Forecast Detection
Solved a demand forecasting problem for Walmart using Random Forest classification and comprehensive data cleaning techniques. Preprocessed Walmart sales history dataset, handled missing values, and engineered features to accurately predict sales forecasts across multiple product categories.
- Python
- Random Forest
- Data Cleaning
- Feature Engineering
- Time Series