Software Developer Los Angeles

Manuj Kumar

213-***-**** **************.**@*****.*** LinkedIn - Manuj Kumar Los Angeles, CA

Professional Summary

Software Developer with 4+ years of software development experience, specializing in scalable web applications and distributed systems. Proven ability to design, build, and maintain enterprise-grade systems using Java, React, Spring Boot, and Microservices. Skilled in agile environments, RESTful API design, object-oriented programming and cloud infrastructure (AWS, GCP). Strong understanding of CI/CD, unit testing, and cross-functional collaboration, with a passion for delivering robust and elegant software solutions. Education

University of Southern California Jan 2024 – Dec 2025 Master of Science in Computer Science GPA: 3.88/4.0 Los Angeles, CA

• Relevant Coursework: Analysis of Algorithms, Databases, Machine Learning for Data Science, Deep Learning, Software Engineering, Natural Language Processing

Kurukshetra University July 2016 – May 2020

Bachelor of Technology in Computer Science and Engineering CGPA: 8.95/10 Kurukshetra, Haryana, India

• Relevant Coursework: Data Structures, Operating Systems, Web Technologies, Object-Oriented Programming, Mobile Computing, Cloud Computing, Distributed Systems Skills

Programming Languages: Java, Python, JavaScript, Golang, SQL, C++, HTML, CSS Frameworks: Spring Boot, Microservices, NestJS, Django, GraphQL, REST APIs, Bootstrap, ReactJS, Angular DevOps & Cloud: Docker, Kubernetes, CI/CD configurations, Firebase, Jenkins, AWS (Lambda, S3), GCP, Azure Databases: MySQL, PostgreSQL, MongoDB, Couchbase, Oracle, Redis Testing: Jest, React Testing Library, Postman, Selenium, Mockito, Junit Data & ML: PyTorch, Tensorflow, LSTM, LLMs, Time Series Analysis, Data Visualization Tools: Git, JIRA, Jenkins, Power BI, Apache Kafka, Elastic Search, Typesense Development Practices: Agile/Scrum, TDD (Test-Driven Development), MVC, SOLID principles Professional Experience

Vosyn Inc May 2025 - Present

Software Engineer Intern Chicago, IL

• Built and maintained scalable RESTful APIs using Django, Django REST Framework (DRF), and FastAPI; integrated backend with PostgreSQL and NoSQL databases for efficient data handling.

• Developed dynamic, user-friendly interfaces using React.js, Material-UI, and implemented global state management with Redux to enhance app performance.

• Collaborated with backend, cloud, and data engineering teams; conducted thorough API testing using Postman, and maintained clear documentation with Notion.

Amdocs Development Center India July 2021 – Dec 2023 Software Developer Gurugram, India

• Worked on a microservices-based telecom e-commerce application using Spring Boot, Couchbase (NoSQL), Elasticsearch for search, and Kafka for messaging.

• Resolved critical product performance issues, particularly in external platform integrations, leading to a 27.5% reduction in customer-reported bugs and enhanced product reliability through API and middleware optimization.

• Led enhancements to improve product scalability and robustness by optimizing system architecture, resulting in a 23.71% reduction in load times and a 28.48% decrease in reported performance-related incidents. Dataguise - A Pkware Company July 2020 – July 2021 Software Engineer Mohali, India

• Worked on data security application that detect and mask customer’s sensitive data, follows MVC framework with Java as programming language, Hibernate as ORM framework to interact with database.

• Designed and implemented robust security solutions with development and testing teams, enhancing system protection and resilience, resulting in a 27% reduction in vulnerabilities.

• Resolved performance bottlenecks and critical bugs, reducing system downtime by 18% through root cause analysis, efficient debugging, and optimization of resource allocation, memory usage, and server response times. Projects

USC Marketplace - [GitHub] Jan 2025 - April 2025

Golang, React, GCP, MongoDB, Jenkins

• Developed a full-stack secure online marketplace for USC students to buy and sell items, using Golang (backend), MongoDB (database), and React (frontend).

• Designed using a Microservices Architecture, comprising distinct modules such as Authentication, Marketplace, Communities, Housing, and others.

• Implemented JWT-based authentication for secure login and user management, designed RESTful APIs for efficient communication between microservices.

Predicting NFL Game Outcomes Using Time-Series Analysis - [Project Report] Aug 2024 - Nov 2024 Python, XGBoost, ARIMA, LSTM, Time-Series Forecasting, Data Visualization

• Collected and engineered team-level NFL data, incorporating Elo ratings and annotated weather data to develop predictive models for game outcomes.

• Implemented and evaluated multiple models including Logistic Regression, XGBoost, MLP, RNN, and LSTM; LSTM achieved the highest accuracy at 74.63%.

• Enhanced the Elo rating system using feature importance from XGBoost and integrated temporal-spatial weather features (e.g., wind speed, temperature, precipitation). Hate Speech Unlearning for LLMs - [Project Report] Jan 2025 - April 2025 LLaMa 2–7B, LLaMa 3.1–8B, LoRA, PyTorch, BLEURT, Contrastive Learning

• Reimplemented and enhanced the Selective Knowledge Unlearning (SKU) methodology to reduce hate speech in LLMs while preserving accuracy on safe prompts.

• Implemented gradient ascent and mismatch-based methods from recent benchmarks to evaluate and establish a strong baseline for unlearning efficacy.

• Integrated a contrastive learning module and a novel Relearning stage to mitigate performance drops on control tasks, improving BLEURT scores and reducing harmful response rates.

• Evaluated model performance on metrics like harmful rate, perplexity, and BLEURT using PKU-SafeRLHF and TruthfulQA datasets.

Awards & Achievements

• Peer 2 Peer Recognition Award (Twice in a Row), Amdocs Development Center India

• Special Achievement Award, Kurukshetra University

Contact this candidate