I am a passionate developer with a solid foundation in VLSI design and a knack for problem-solving. With 1.5 years of experience at Infosys as a Systems Engineer, I have honed my skills in developing efficient software solutions and collaborating with diverse teams. My education at Amrita University provided me with a strong technical background. During my college years, I completed a project in digital beamforming for 5G sequences on a Basys 3 FPGA Board using Verilog coding, which gave me hands-on experience in hardware design and implementation. I further enriched my skills through an internship at Continental Automotive. There, I contributed to a project on synthetic data generation using Carla and Scenic with Python, showcasing my ability to apply theoretical knowledge to practical challenges. I thrive in dynamic environments and am driven by a commitment to innovation and continuous learning.
| Name | Description |
|---|---|
| FPGA implementation of Digital beamforming | Digitalization has increased signal interferences, necessitating effective signal separation. Digital beamforming with the Least Mean Square (LMS) algorithm is promising but suffers from delays, overshooting, and instability due to its constant step size, especially with varying Signal to Noise Ratios (SNR). To improve this, a Variable Step Size (VSS) LMS algorithm is used. Through MATLAB simulations it was found that VSS LMS outperforms Constant Step Size (CSS) LMS in error convergence, beam pattern quality, and performance under varying SNRs. The architecture was designed and simulated on a Basys 3 FPGA board using Verilog, with resource utilization analyzed in Vivado software. |
| Synthetic data generation for autonomous driving | Synthetic data was generated for a parking scenario by integrating SCENIC and CARLA software with Python. Various parking scenarios were tested across different CARLA towns and maps. Data and annotations were collected using sensors such as cameras, radar, and lidar, capturing 2D bounding boxes, semantic segmentation, and depth imaging. Finally, a user interface for automatic synthetic data generation was created using a Streamlit app. |
| SPI(Serial Peripheral Interface) protocol implementation using Xilinx Vivado | The varying temperature in a room was recorded using an ambient light sensor and displayed on an LED screen via a Basys 3 FPGA board using Verilog coding. The SPI protocol connected the Basys 3 board to the ambient light sensor. Temperature values were continuously read and displayed on a scale of 1 to 255. |
| Sales prediction using machine learning algorithm | A predictive model was built to determine the sales of each product at Walmart stores. This model helped identify key properties of products and store attributes that significantly influence sales performance. |
| Intelligent waste segregation system | Improper waste management in India causes health and environmental issues. Developed a recycling Machine, using a Microsoft webcam and image processing. This device is efficient in reducing errors, and minimizes human labor, improving traditional waste management methods. |
- Apply VLSI design, synthetic data, and machine learning expertise to help Kerala startups innovate.
- Focus on sustainable tech solutions like intelligent waste management.
- A highly literate and technically skilled workforce is a significant asset.
- Enhance collaboration between academic institutions and industry to foster research and innovation.
- Promote sustainable and eco-friendly technologies to address environmental challenges, leveraging Kerala’s natural resources and eco-conscious population.