• Developed a customizable AI-powered Applicant Tracking System (ATS) using Python and libraires such as Pandas and scikit-learn that ranked resumes based on role-specific qualifications, increasing resume screening speed by 70% and improving candidate prioritization for the TMU Intramural Official role
• Upgrading and improving website functionality using Adobe Experience Manager (AEM), increasing system reliability by 40% and reducing content deployment issues by 30%
• Organized records and documentation, improving retrieval by 60%, ensuring compliance, and exploring blockchain methods to secure data

Developed and deployed a Python-based validation testbench for sensor interfacing, automating over 95% of data
verification tasks for chipless RFID sensor simulations and increasing consistency across SPI/I2C interfaces
• Designed a multi-layer PCB schematic for degradation testing, enabling accurate signal routing and component layout,
which supported a 40% reduction in prototyping errors during fabrication and initial testing phases
• Created a PCB testbench draft using UVM and integrated it into the existing QA pipeline, accelerating debugging by 50%
enhancing test coverage for environmental durability scenarios, and supporting production validation

Authored a comprehensive 10-page technical paper utilizing Word to enhance the integration of CISCO-Meraki
technologies within public transit systems, contributing to an estimated 25% increase in operational efficiency
• Engineered and deployed a test MAN using PuTTY, contributing 100+ lines of code, which improved server operation
efficiency by 30% and enhanced cybersecurity protocols by reducing vulnerabilities by 15%.
• Accelerated stakeholder endorsement by 30% by designing and developing a 19-slide presentation pitch that showcased the integration of IoT and LiDAR sensor technologies within smart infrastructure solutions

• Developed a robust Node.js and Express REST API integrating MongoDB, JWT authentication, and blockchain for securing keys and data, supporting ~250 receipt analysis requests per day during testing.
• Designed and implemented a TensorFlow-based model service capable of detecting fraudulent receipts with 92% classification accuracy across more than 20 training samples across 50 epochs by analyzing audit values
• Integrated the Gemini API to extract structured receipt metadata, reducing manual data entry time by approximately 70% by processing image buffers and returning merchant names, transaction amounts, and dates
• Created responsive React user interfaces styled with Tailwind CSS, including drag-and-drop receipt uploads and dynamic dashboards, which improved user engagement by 45% as measured through simulated analytics during testing

• Building an autonomous RC car with TensorFlow, OpenCV, and Verilog troubleshooting scripts on Raspberry Pi, achieving 85% tracking accuracy and cutting sensor-related downtime by 40%
• Implementing FPGA logic using VHDL for power regulation features, aiming for a 60% improvement in PWM timing

The Student Grade Predictor utilizes randomly generated student scores to predict whether a student will fail or pass based on their grades and Attendance. This project includes a sample dataset to train the network; however, the project automatically creates a new dataset every instance run to train the data set.

Designed and implemented a 4-bit ALU in Verilog to perform arithmetic and logical operations (such as addition, subtraction, AND, OR, XOR, NOT, and bitwise shifts) on 4-bit inputs based on a 3-bit control signal, successfully producing a 4-bit output with status flags (Zero, Carry, Overflow, Negative) to support fundamental CPU operations.

• Developed a team working to designing a real-time productivity tracking application with a PyQt5 interface and machine learning models to classify user activity as productive or unproductive, improving time management by 30%
• Implemented AI-powered reminders leveraging scikit-learn to deliver targeted prompts, increasing user efficiency by 25%
• Automated system usage classification through QML integration and continuous model training, reducing manual tracking and achieving 90% classification accuracy

The Wireless Robotic Hand project aims to develop an assistive device designed to help hospital patients suffering from nerve damage or hand mobility issues. Utilizing flex sensors, Arduino programming, and wireless communication, the robotic hand will mimic the movements of a healthy hand, providing patients with a means to perform everyday tasks with greater ease and independence.