Software Engineer United States

HARESH INDRAJIT

Seattle, WA, United States +1-206-***-****

******@**.*** www.linkedin.com/in/haresh-indrajit EDUCATION

University of Washington - Master of Science in Electrical Engineering, Seattle, WA, United States Mar 2025 Coursework: AI for Mobile Robots, Large Language Models (LLMs), Computer Vision, Data Structures and Algorithms Anna University - Bachelor of Engineering, Chennai, India Aug 2021 Coursework: Python Programming, Cloud Computing, Machine Learning, Video Analytics TECHNICAL SKILLS

● Frameworks/tools: React, Django, NextJS, NodeJS, Angular, React Native, Tensorflow, Pytorch, Spring Boot, Flask, mySQL, GraphQL, PostgreSQL, Docker, Kubernetes, Selenium, MongoDB, Git, Github, sXCode, Android Studio, Flutter, Hadoop, LangChain, CUDA, ROS 2, TensorRT-LLM, cuDNN, Stripe, GraphQL, WebRTC, Tableau, Simulink, JIRA, Unity3D, Jenkins, Open vSwitch, Supabase, NCCL, JAX, SWIG, ctypes, Cython, Linux Operating Systems

● Programming: Python, Javascript, TypeScript, Rust, C, C++, Java, C#, HTML, CSS, Swift, SQL, MATLAB, Verilog, bash, AJAX

● Cloud Infrastructure: AWS (ECS, EC2, EBS, AMIs, S3, RDS, QuickSight, Elastic Kubernetes Service - EKS) EXPERIENCE

JumpStartCSR, Seattle, Washington - Software Engineer Intern Jun 2024 - Sept 2024

● Developed a motion-tracking solution for cavalry personnel to assess movement while wearing body armor in unrestricted environments, through a both a React/Django web application and a iOS mobile app in Swift using Apple’s Vision API for real-time, accurate motion tracking on mobile devices, accessible in field conditions, achieving a processing speed of 30 frames per second with 90% accuracy in movement tracking, providing a reliable, dual-platform solution for field operatives ToCode, Chennai, India - Software Engineer Aug 2021 – Jan 2023

● Developed a web application that could perform fine-grained object detection and text recognition on ID cards to automate and streamline the verification process for the sale of financial products, by building a React front end with a Django backend for the web app and React Native cross-platform for the mobile app to capture ID card images, using PyTorch to integrate state-of-the-art detection models, leveraging PANet for text detection and RobustScanner for text recognition, achieving an mAP of 0.95 for object detection and an 88% accuracy rate in text detection HippoVideo, Chennai, India - Software Engineer Intern Dec 2020 – Mar 2021

● Developed a web module that could enhance video resolution, improving visual quality for users uploading low-resolution content, using Angular on the front end and Django on the backend, deploying the solution on AWS EC2, using a Knowledge Distilled Enhanced Super Resolution GAN (ESRGAN) model to enhance videos and optimize processing for high efficiency, achieving a Peak Signal-to-Noise Ratio (PSNR) of 29.79, allowing users to upload and enhance videos directly through the web app for clearer, high-quality playback

PROJECTS

University of Washington(Seattle) Sponsored Industry Project Jan 2024 - May 2024

● Developed a web application to help ecologists efficiently annotate and record biodiversity anomalies in specific areas, using ReactJS for the front end, NodeJS and FastAPI for the backend, and integrated three.js, leaflet.js, and tensorflow.js for interactive mapping and image analysis, accommodating up to 5 image uploads and 200 annotators simultaneously completing photogrammetry in under 2 hours

ML Anomaly Detection System for Enhanced Threat Detection Mar 2024 - May 2024

● Developed a machine learning-based anomaly detection system for AWS through a scalable Python backend using CloudTrail and GuardDuty logs by implementing unsupervised learning algorithms, including Isolation Forest and DBSCAN, to detect unusual patterns in API usage and network traffic, enabling real-time identification of advanced persistent threats (APTs) and insider attacks. achieving a 50% improvement in detection coverage, surpassing the limitations of traditional rule-based systems. Multi-Agent Chatbot with a Predefined Source of Truth Feb 2024 - Mar 2024

● Developed a question-answering chatbot system that used a specific book as the authoritative source to provide accurate responses aimed at reducing reliance on costly API queries, using ReactJS for the frontend, Django for the backend, Pinecone for vector storage, OpenAI's GPT-3.5 Turbo for language processing, and MySQL for logging interactions implementing optimization techniques to reduce API costs by 40%, with the chatbot accurately answering 95% of queries, verified through qualitative assessments

Bachelor's thesis - ParvAI - A Smart Eyes System for the Visually-Impaired Sept 2020 - Mar 2021

● Developed a deep learning-based assistance system to enhance navigation for visually-impaired individuals, using a dual-component system combining Semantic Segmentation (DeeplabV3) for path identification and a Transformer-based Image Captioning model in PyTorch for generating scene descriptions to process and relay information at real-time speeds, achieving a real-time processing speed of 15 frames per second, improving users’ environmental awareness and navigation capability by 25%, based on initial user feedback

Contact this candidate