Dhushyanth Manjunath
Software Engineer — Systems
CURRENT
Software Engineer at The Soco App Corp.
PAST
Former Training Assistant at UT Arlington Police Department
Featured Projects.
AI Documentation Generator
A lightweight tool that leverages open-source LLMs to automatically generate documentation and unit tests for Python code using local models via Ollama.
PDF Question Answering System
A production-ready Retrieval-Augmented Generation system for querying PDF documents using local AI models with PostgreSQL pgvector for semantic search.
Packet Processor
A high-performance packet processing engine that captures, analyzes, and processes network packets in real-time with zero-copy buffer optimization and efficient protocol parsing.
Object Detection System
A YOLOv8-based object detection system deployed as a FastAPI web service for identifying pedestrians and vehicles in images with Docker containerization.
VRP Optimizer
A sophisticated logistics optimization system implementing the Adaptive Large Neighborhood Search algorithm to solve Vehicle Routing Problems with real-world constraints.
Elevator Simulator
A multi-threaded elevator system simulation with advanced scheduling algorithms optimizing passenger flow in high-rise buildings.
Multi-threaded Image Search Engine
A secure and efficient multi-threaded image search engine implemented using gRPC, Docker, Python, and Protocol Buffers, featuring advanced image retrieval algorithms.
Smart Traffic Sign Detection
Built a traffic sign recognition system using TensorFlow and OpenCV, leveraging computer vision and cloud-based deployment for real-time detection.
Spark AWS Data Pipeline
A cloud-scale data processing pipeline following best practices with cost optimization through Spot Instances, efficient Parquet storage, and auto-termination of resources.
Mac Framer
A modern web application that transforms screenshots and images into elegant MacBook-style mockups with just a few clicks, built with TypeScript.
Audio Transcriber
Audio transcription tool using advanced speech recognition APIs to accurately convert spoken language into text with support for multiple languages.