I am an undergraduate Computer Engineering student at Kathmandu University, Nepal, working across Full-Stack Web Development and Applied AI Systems. My technical focus spans building scalable backend REST APIs, responsive frontend web interfaces, and deploying computer vision and deep learning models into production environments.
class Pranaya:
def __init__(self):
self.name = "Pranaya Shrestha"
self.education = "Computer Engineering at Kathmandu University"
self.location = "Nepal"
self.specializations = [
"Computer Vision & Deep Learning",
"Full-Stack Web Development",
"REST API Design & Backend Architecture",
"Applied AI Systems Integration"
]
self.stack = {
"ai_ml": ["PyTorch", "OpenCV", "YOLO-NAS", "HuggingFace"],
"backend": ["FastAPI", "Python", "Node.js", "REST APIs"],
"frontend": ["React", "JavaScript", "HTML5", "CSS3"],
"database_tools": ["MySQL", "Git", "GitHub Actions", "Docker"]
}
self.looking_for = "SWE / ML Internships & Entry-Level Roles"
def say_hi(self):
return "Let's build something scalable and impactful together."- Backend Development: Designing scalable REST APIs, microservice architectures, and data processing layers using FastAPI, Python, and Node.js.
- Frontend Engineering: Creating responsive, component-driven user interfaces using React, JavaScript, HTML5, and CSS.
- Computer Vision & AI: Developing depth-estimation, object detection, gesture recognition, and bioacoustic models with PyTorch, YOLO-NAS, and OpenCV.
- Systems & Automation: Connecting web applications to AI pipelines, background job queues, and automated media processing workflows (FFmpeg, ComfyUI, Edge-TTS).
|
Geometry-agnostic detection & depth pipeline Team project combining YOLO-NAS-S object detection with monocular depth estimation and synthetic data generation to identify retail items regardless of shape or packaging. Key Contributions: Built shape classification algorithms, evaluated depth extraction models (YOLOv8-Seg, Detectron2, Depth Anything V2), and conducted error analysis.
|
Full-stack bioacoustic monitoring · Winner — EcoThon Innovation Track Award-winning full-stack hackathon project scoring forest health quality from audio recordings. Integrates BirdNET species detection with ecological datasets through a custom-trained MLP backend. Key Contributions: Built the
|
|
Conversational chatbot & full-stack web API A supportive conversational chatbot application for students navigating stress, isolation, or anxiety — providing accessible, low-friction mental health support through structured APIs and natural dialogue.
|
Automated end-to-end video production system Automated short-form content engine chaining diffusion-based image generation (ComfyUI), Edge-TTS narration, and FFmpeg assembly through asynchronous Python processing.
|
- ASL Sign Language Recognition: Real-time gesture recognition and computer vision translation system using OpenCV and PyTorch. github.com/Pranaya-sht/ASL
Languages
Backend & API Engineering
Frontend Engineering
AI / Deep Learning / Computer Vision
Databases & Developer Tools
GitHub Trophies
I am actively seeking Software Engineering, Full-Stack, and Machine Learning Internships or Entry-Level opportunities. Feel free to reach out directly through any of the channels below:
