Software Engineer
Safal Raj
Manandhar
Building full-stack applications and cloud-native solutions with Python, TypeScript, and AWS. Passionate about turning ideas into reliable, scalable software.
Software Engineer
Building full-stack applications and cloud-native solutions with Python, TypeScript, and AWS. Passionate about turning ideas into reliable, scalable software.
A bit about my background
I'm an Associate Software Engineer at Maitri Holdings Services, where I develop and maintain software using Python, TypeScript, and deploy on AWS. I graduated with a B.E. in Computer Science from Khwopa College of Engineering, Tribhuvan University.
I've published research on Paddy Leaf Disease Detection using CNN-LSTM achieving 95% accuracy, and hold an AWS Certified Cloud Practitioner certification. I enjoy building across the full stack — from clean APIs to responsive interfaces to cloud infrastructure.
1+
Year Experience
5+
Projects
1
Research Paper
Technologies I work with
Where I've worked
Maitri Holdings Services — Jhamsikhel, Lalitpur
Leapfrog Technology — Dillibazar, Kathmandu
Things I've built
Connecting Farmers and Customers Directly
Full-stack web application connecting farmers directly with customers. Node.js/Express.js backend with vanilla TypeScript frontend.
Browser-Based Survival Game
Survival game where players battle waves of zombies. Built with Canvas API and TypeScript during the Leapfrog fellowship.
AR Navigation with Voice Assistance
AR navigation to monuments in Bhaktapur Durbar Square using Mapbox with voice assistance. Uses MobileNet-SSD v2, CNN-LSTM, Transformers, and Whisper AI.
Rice Disease Detection — CNN-LSTM
Image classification to detect paddy leaf diseases with 95.50% accuracy. Built with TensorFlow/Keras, deployed as Flutter mobile app on Azure VM.
Pure NumPy Implementation
Neural network for regression and classification built entirely from scratch using NumPy. Based on Calculus for ML & Data Science course.
Published Research
Combining features extracted by VGG16 and VGG19 on segmented images, fed to an LSTM for classification — achieving 95% accuracy. Generalizable to other image classification domains.
Academic background & credentials
Khwopa College of Engineering, Tribhuvan University
Libali, Bhaktapur, Nepal — April 2024
AWS Certified Cloud Practitioner
Amazon Web Services
SAP S/4HANA Cloud Public Edition, Sourcing & Procurement
SAP
Calculus for Machine Learning and Data Science
Coursera
Scientific Computing & Python for Data Science
World Quant University
Data Visualization
Kaggle
Stories and experiences beyond code
Let's connect
I'm open to new opportunities and interesting projects. Feel free to reach out.
