ROSHWIN RUZAR FERNANDES

Software Quality Assurance:

• Oversaw automated test environments, ensuring software quality compliance with Microsoft test suites (System HLK) in 10+ Domains.
• Diagnosed and isolated system failures, identifying root causes and working collaboratively with internal and external stakeholders across 7+ domains, including Power Management, Connectivity, Graphics, Imaging, Storage, Firmware and Audio.
• Experienced in Level 1 debugging, utilizing tools such as WinDbg and ITP to address both component-level and system-level issues. Proficient in diagnosing and resolving critical problems, including BSOD (Blue Screen of Death) and TDR (Timeout Detection and Recovery) issues.

Test Management & Execution:

• Managed automated test environments and collaborated with hardware, firmware, and software teams to identify root causes critical issue resolution in collaboration with the Microsoft OS team.
• Interfaced with hardware, power, thermal, and software engineers to understand system architecture and execute test suites.
• Contributed to meeting quality release criteria and platform exit milestones.

Testing Expertise:

• Mentored junior engineers in best practices for validation engineering methodologies improving team skillsets, coordinated milestone execution, and drove issue resolution to meet deadlines.
• Providing detailed validation data and debug analysis, removing project milestone blockers, and building strong internal and external relationships.

Hands-on experience with a wide array of tools and devices, including:

• ETL, ITP, XDP, Socwatch, Pythonsv, Gpuview, GDHM, Debugview, WinDbg, PowerHouseMountain, NPK, WRT, WireShark, Matthew Audio, Ellisys, Firmware Bios and the Windows Performance Analyzer (WPA).

GPA: 7.6 CGPA

Capstone Project - Collaborative filtering Recommendation system using K- Nearest neighbours Algorithm for Books dataset

• Developed a project focused on data visualization and exploratory data analysis to better understand user preferences.
• Implemented a recommendation system using the K-Nearest Neighbor (KNN) algorithm to suggest books from a large database.

Capstone Project - Classifying News Article as Fake or Legitimate using Artificial Neural Network

• Implemented a Deep Neural Network (DNN) model to predict outcomes based on input variables derived from news text.
• Utilized Natural Language Processing (NLP) techniques to extract key textual features and determine optimal weight selection for the model.