Hi, I'm Nirmal
Data Scientist & Analytics Professional
I transform complex data into actionable insights that drive business decisions. With expertise in machine learning, statistical analysis, and data visualization, I help organizations unlock the power of their data.
View My Work
About Me
I am a Data Scientist with a software-engineering backbone and proven experience delivering end-to-end AI and data-driven solutions - from data wrangling and feature engineering to modeling, evaluation, and scalable production deployment.
In prior roles, I modernized enterprise platforms in regulated settings - migrating legacy systems to a modern Python microservices architecture, designing secure APIs for core transactional workflows, executing large-scale data migrations with minimal downtime, and strengthening role-based authentication, encryption, and auditability. I also delivered operational data applications on Linux/cloud infrastructure and designed data pipelines that supported analytics for performance, risk, and resource planning. Across these projects, I introduced CI/CD and observability and shipped dashboards that turned operations into decision-ready insights while laying a foundation for ML.
To enhanced my predictive analytics skills, I completed Advanced Data Analytics Professional Certificates with real-time analytics and machine learning implementation, and I also earned Microsoft Azure AI Fundamentals (AI-900) to ground cloud-AI concepts.
Besides technical work, I take part in AI communities, mentor aspiring Python developers, and contribute to open-source projects - practices that keep me collaborative, innovative, and continuously learning.
Technical Skills
Programming & Frameworks
- Python
- Scikit-learn
- NumPy
- Pandas
- Statsmodels
- PyTorch
- Hugging Face
- LangChain
- FastAPI / Django
ML & Statistical Modeling
- Regression
- Classification
- Clustering
- Tree-Based Models
- A/B Testing
- Hypothesis Testing
- Time Series
Data Visualization
- Power BI
- Tableau
- Matplotlib
- Seaborn
- Plotly
- EDA (Exploratory Data Analysis)
ML Techniques
- LLMs
- LLM Finetuning
- Hugging Face
- RAG
- Embedding Models
- Vector DB(FAISS)
- Azure OpenAI
- Azure Cognitive Services
- BERT
Tools & Deployment
- Azure / AWS
- Git / GitHub/ Gitlab
- SQL / PostgreSQL
- Apache AirFlow / MLFlow
- MLOps
- Docker / Kubernetes
- VPS (Ubuntu) / Nginx / Gunicorn
My Projects
Employee Churn Prediction
Built an end-to-end churn prediction system with scikit-learn pipelines and a Streamlit app, achieving 96.2% accuracy and delivering real-time risk scores with actionable retention insights.
Fine Tuning Transformers BERT & BioBERT
Fine-tuned BERT & BioBERT on the BC5CDR dataset to extract chemical and disease entities from PubMed abstracts, boosting F1-score by 27%, accuracy by 9%, and optimizing precision by 28% and recall by 26% over baseline.
Mental Health Prediction
Built an ML pipeline to predict mental health risks from social media and demographic data. Achieved 71.7% accuracy with tuned Logistic Regression, outperforming more complex models, and delivered insights on model interpretability and dataset limitations.
Movie Review Sentiment Predictive Analysis
Implemented sentiment predictive analysis using TF-IDF and Word2Vec embeddings, applying XGBoost and Random Forest classifiers, and achieved 83% accuracy.
Automobile Prediction Analysis
Automobile Price Prediction – Built a machine learning pipeline to estimate vehicle pricing based on design, performance, market, and temporal attributes. Hyperparameter tuning improved R² from 95.6% -> 96.5% and reduced MSE/RMSE, making the Decision Tree model more robust and generalizable.
Main Certifications
Microsoft Azure AI Fundamentals (AI-900)
Credential validating foundational knowledge of AI concepts and Azure Cognitive Services.
Google Advanced Data Analytics Professional Certificate
~160 hours of hands-on projects covering regression, classification,clustering,A/B testing, Hypothesis testing, ML pipelines, and real-world analytics case studies.
Achievements & Honors
Hackathon Winner
Brainycube Research OrganizationWon first place in a 48-hour hackathon by building a performant backend API for a dating app.
Virinchi College & Brainycube, KathmanduNational Robotics League (with International Delegates)
Yantra 7.0Lead a team of 5 for 2.5 months to design football-playing robots; won Best Design among national & international teams.
Robotics Association of NepalDean's Honours
Fall 2024 & Winter 2025Recognized for academic excellence in the Artificial Intelligence & Machine Learning program.
Lambton CollegeInnovation Showcase – ICT Mela 1.0
ICT Mela 1.0Showcased Smart Home IoT and a Robotic Football Car; praised for creativity and execution.
Virinchi College & RAN