Hi there! 👋 I am Shahrukh Ahmad, an ambitious Computer Science student, developer, and continuous learner. I believe the best way to master technology is by rolling up your sleeves and building real-world, production-ready systems from the ground up.
My technical interests and day-to-day focus revolve around:
- 🌐 Modern Web Development & Distributed Streaming: Learning how real-time network streams work, managing in-memory byte buffers with Node.js and Express, and reducing server storage overhead.
- 👁️ Computer Vision & Deep Learning: Training custom object detection and instance segmentation models using PyTorch, YOLOv8, and OpenCV.
- 🔒 Privacy-First Client Engineering: Building offline-capable Progressive Web Apps (PWAs) using HTML5 Canvas, SVG vectors, and mathematical error-correction (Reed-Solomon algorithms).
- ☁️ DevOps & Automated Workflows: Exploring CI/CD automation pipelines using Google Cloud Platform (GCP Cloud Build) and Git.
| Project | What I Built & Learned | Live Platform & Code |
|---|---|---|
| FB4K-StreamEngine | Media Streaming & In-Memory Transcoding Pipeline • Built a high-speed web utility for public media archiving. • Learned socket pooling, chunked transfer encoding, and in-memory piping. • Integrated FFmpeg for audio demuxing (320kbps MP3) and resolution upscaling. • Implemented statutory DMCA and GDPR cookie consent standards. |
🌐 Live Website 💻 Source Code |
| MyLogoQR | Client-Side Custom QR Generator & PWA • Engineered a 100% browser-based tool with zero server data storage. • Implemented Reed-Solomon Level H fault tolerance (30% error correction). • Enabled offline functionality using Service Workers and Cache Storage. • Designed custom gradients, logo compositing, and SVG vector exports. |
🌐 Live Website 💻 Source Code |
| YOLOv8 Cow Segmentation | Agricultural Deep Learning & Instance Segmentation • Curated custom datasets and performed detailed polygonal mask annotations. • Trained YOLOv8 segmentation models using PyTorch and evaluated mAP scores. • Built an inference pipeline for practical livestock identification and tracking. |
💻 Source Code |
| Speech-to-Text Whisper | Local Transformer Speech Recognition Application • Built a desktop audio transcription utility using OpenAI's Whisper model. • Handled multi-lingual audio processing locally without external paid APIs. |
💻 Source Code |
Languages : Python, JavaScript (ES6+), HTML5, CSS3, SQL, Shell
Frameworks & ML : PyTorch, YOLOv8, OpenCV, Express.js, Node.js, Flask
Tools & DevOps : Git, GitHub Actions, GCP Cloud Build, FFmpeg, Docker, Apache
Core Concepts : Asynchronous Programming, In-Memory Streams, Computer Vision, PWA
I am always eager to learn, contribute to open-source, and connect with fellow developers, mentors, and professors.
- 📧 Email: srkamduk@gmail.com
- 💼 LinkedIn: linkedin.com/in/srkamd
- 🌐 Projects: fb4kdownloader.com • mylogoqr.com