MANNSHRESTHA©
Illustrative project concept
- Contract
- Validate
- Quarantine
- Publish
Building intelligent systems
from research to deployment.
02 / Selected work
Data with
a purpose
(10 projects)
A look inside the systems. Illustrative concepts are labeled on each project.
03 / The person behind the pipelines
I’m a Machine Learning Engineer focused on computer vision, predictive modelling, and bringing models into useful applications. My experience spans ML engineering at Mantrain Nepal and data science at SiliconTech Nepal, from preprocessing and model evaluation to deployment. My academic background includes computer engineering and postgraduate study in Artificial Intelligence at the University of East London.
Computer vision
Detection & segmentation
Python + PyTorch
Models & applications
MLOps
Reproducible delivery
Experience.
Feb 2024 — Sep 2025
Machine Learning Engineer
Mantrain Nepal Pvt. Ltd.
Developed tumour segmentation systems and evaluated detection and segmentation architectures, including U-Net, Link-Net, YOLOv8, and Detectron2. Researched transfer learning for domain-specific datasets.
Aug 2022 — Nov 2023
Data Scientist
SiliconTech Nepal Pvt. Ltd.
Worked on a recommendation engine, model validation, and computer vision solutions across detection, segmentation, tracking, classification, and depth estimation.
Jan 2022 — Aug 2022
Data Scientist Intern
SiliconTech Nepal Pvt. Ltd.
Built preprocessing pipelines and automated reporting with Python, Pandas, and NumPy. Developed model-backed applications with Flask and Streamlit.
Education.
Sep 2025 — Sep 2026
MSc Artificial Intelligence
University of East London
Study listed in the supplied résumé: deep learning, computer vision, natural language processing, scalable ML systems, and AI ethics.
Oct 2016 — Jun 2021
B.E. Computer Engineering
Cosmos College of Management and Technology
GPA 3.48 / 4.00. Coursework included data structures and algorithms, machine learning, database systems, computer networks, and OOP.
Training & certifications.
- Python with Artificial Intelligence (AI) Training
- MATLAB in Computer Vision
04 / Areas of focus
What I
build for
- (01)
Computer Vision
Object detection, instance and semantic segmentation, tracking, classification, and depth estimation with modern deep learning tools.
PyTorch · TensorFlow · YOLOv8 · Detectron2 · OpenCV
- (02)
Machine Learning
Predictive modelling, preprocessing, feature engineering, cross-validation, and model evaluation with reproducible experiments.
Python · Scikit-learn · XGBoost · Pandas · NumPy
- (03)
MLOps & Deployment
Versioned data and experiments, modular training pipelines, and model-backed web applications.
MLflow · DVC · DagsHub · Docker · Flask · Streamlit
- (04)
Language & Retrieval
Retrieval-augmented applications that connect curated knowledge with language models and a usable interface.
LangChain · RAG · Hugging Face · FastAPI