SabinLamsal
AI/ML Engineer crafting production-ready intelligent systems — from detecting crop diseases at 98.18% accuracy across 45 diseases to deploying recommendation engines trained on 1M+ ratings.
class MLEngineer: name = "Sabin Lamsal" location = "Kathmandu, Nepal" focus = [ "Computer Vision", "Time-Series Forecasting", "Recommendation Systems", "NLP" ] stack = ["TensorFlow", "Scikit-learn", "XGBoost", "Streamlit"] accuracy = 98.18 # crop CNN, 45 diseases projects = 42 # github.com/sabin74 open_to = True # available now 🟢
Building AI that
solves real problems
I'm an aspiring Machine Learning Engineer pursuing my Bachelor's in IT/CS at Tribhuvan University (Currently Running). My work spans agriculture, finance, retail, and entertainment — anywhere data can drive smarter decisions.
During my internship at Miss Misoo Production, I engineered recommendation engines using SVD and session-based models, achieving a 20%+ accuracy improvement over baseline and deployed two real-time Streamlit web applications.
My flagship project — an Agriculture Crop Disease Detection & Advisory System — achieves 98.18% accuracy across 45 diseases on 14 crops, using MobileNetV2 transfer learning in a bilingual MERN-stack deployment tailored for Nepali farmers.
Professional Background
- Engineered recommendation engines using popularity-based, SVD, and session-based models — achieved >20% accuracy improvement over baseline.
- Solved cold-start and 95% sparsity challenges through hybrid and session-based recommendation approaches.
- Built and deployed two real-time Streamlit web applications for movie and retail product recommendations.
- Built end-to-end ML/DL pipelines across 42+ public GitHub repositories.
- Developed supervised/unsupervised models and deep learning architectures (ANN, CNN, RNN, LSTM, GRU) with TensorFlow/Keras.
- Applied transfer learning (MobileNetV2, VGG16) for computer vision and Hugging Face for NLP.
- Performed comprehensive EDA, preprocessing, and statistical analysis across diverse datasets (IPL, Airbnb, COVID-19, Netflix, etc.).
- Created visualizations with Matplotlib, Seaborn, Plotly, and Folium.
42+ GitHub Projects
End-to-end ML/DL systems across computer vision, NLP, time-series, recommendation, and EDA. Click any card to open the repository.
Academic Background
Credentials & Learning
Let's build
something together
Open to internships, entry-level ML/AI roles, and research collaborations — on-site in Kathmandu, hybrid, or remote worldwide. Immediate start available.
Open for new opportunities
Whether you have a dataset that needs making sense of, an ML system to build, or just want to discuss AI/ML — I'd love to connect.