
1
1
India
1 year of experience
I am Abhishek J S, a Software Development Engineer with a background in Computer Science and Engineering. I have hands-on experience in both software and embedded systems, with a strong focus on C/C++, backend development, and system-level programming. I began my professional journey with a one-year apprenticeship in Qt development, where I built desktop applications and worked with multimedia tools such as FFmpeg. Since then, I have developed multiple projects, including a NES emulator using Qt and C++, a screen recording application, and a feature-rich Tetris game using SFML. I also have practical experience in embedded systems and digital design, working with technologies such as VHDL, Verilog, and FPGA-based systems. My knowledge extends to high-speed interfaces like DDR memory, Ethernet, SPI, UART, and I2C, along with tools like KiCad and others. I am particularly interested in backend development, system design, and low-level programming. Currently, I am preparing for PGCET 2026 while continuing to strengthen my skills in System Architecture, Computer Science and Engineering, and related domains.

Vision-Link AI Hub: Revolutionizing Rural Diagnostics The Problem: In many rural areas across the Global South, access to advanced molecular diagnostics is limited by poor internet connectivity and high costs. Medical practitioners lack the tools to interpret complex CRISPR and genomic data locally. Our Solution: Vision-Link AI is a lightweight, offline-first AI agent designed for the AMD AI ecosystem. It enables healthcare workers to perform rapid molecular diagnostic assistance without needing a cloud connection. By utilizing optimized Small Language Models (SLMs), we bring the power of bioinformatics to the edge. Key Features: Offline Inference: Runs locally on AMD hardware using ROCm, ensuring data privacy and accessibility. CRISPR Analysis: Specialized logic to assist in interpreting molecular diagnostic results. Multilingual Support: Designed to serve diverse global populations. Technical Stack: Built with Python and Next.js, leveraging the Qwen2.5-0.5B model for its high efficiency-to-size ratio. Optimized for AMD accelerators to ensure low latency in critical medical scenarios.
10 May 2026