Hi, I'm Nishan Duwal — a Bachelor of Information Management (BIM) graduate with hands-on experience in Python, Pandas, Scikit-learn, and Machine Learning from my Data Science Internship at TechAxis. I bridge end-to-end data analytics and predictive modeling (SHAP, Naive Bayes, web scraping) with executive storytelling in Tableau & Power BI, backed by full-stack software engineering.
Bridging computational machine learning, exploratory analysis, and business storytelling.
I am a driven Data Science Intern and Software Developer with a Bachelor of Information Management (BIM) from Tribhuvan University. My core expertise sits at the intersection of data engineering, predictive modeling, and executive visual intelligence.
During my intensive internship at TechAxis, I architected end-to-end data pipelines: scraping web data with Playwright, building NLP classifiers with TF-IDF & Naive Bayes, and interpreting feature impacts using SHAP (SHapley Additive exPlanations). I take pride in translating complex algorithmic conclusions into crystal-clear dashboards in Tableau and Power BI.
A balanced toolkit designed to build data pipelines from scratch and turn findings into high-visibility visual products.
Skilled in building predictive pipelines with Naive Bayes, TF-IDF sentiment scoring, and model interpretability with SHAP (SHapley Additive exPlanations).
Transforming raw tabular metrics into intuitive, executive-ready dashboards in Tableau Desktop, Tableau Public, and Power BI with dynamic filters and drill-downs.
Designing relational databases (MySQL, SQLite), backend logic in Flask/Python, AES data encryption security, user authentication, and responsive interfaces.
Automating data harvesting from dynamic web portals using Playwright, Selenium, and BeautifulSoup into structured relational datasets ready for analysis.
An automated, end-to-end financial data product built for tracking the Nepal Stock Exchange (NEPSE). Collects historical time-series data using Python scraping scripts, models it into relational structures, and delivers interactive visual analysis on daily index movements, sector volume surges, and market liquidity trends.
Real-world data science pipelines, machine learning models, BI dashboards, and software applications. Click any image for full-screen inspection.
End-to-end pipeline scraping customer reviews with Python and Playwright. Applied TF-IDF vectorization and Naive Bayes for sentiment classification, then used SHAP for explainable AI to uncover key features driving positive ratings.
Personal Project
Automated Python scraping scripts to collect and structure historical stock market time-series data into relational tables. Built interactive Tableau and Power BI dashboards to track daily price movements, trading volume, and market trends.
Extracted, cleaned, and analyzed 100,000+ job-market records using SQL and Python (Pandas). Conducted rigorous EDA, missing-value treatment, and visual reporting to identify factors behind interview success and salary trends.
BIM Final Project
Full-stack web application built with Python Flask and SQL. Designed database schemas, user authentication, role-based access control, and implemented AES encryption to secure sensitive patient healthcare data.
Web & Analytics
Production-ready e-commerce store with structured product management, secure payment gateways, Google Analytics funnel tracking, and Judge.me customer review integrations for social trust and conversion optimization.
Dedicated repository of algorithmic problem solving, computational complexity benchmarks, and core Python implementations covering dynamic programming, data structures, graph theory, and mathematical optimization.
Categorized breakdown of programming languages, analytics toolsets, machine learning libraries, and frameworks.
Verified professional internships, academic achievements, and industry certifications.
Interested in data science pipelines, BI dashboards, or software development? Send a message or reach out directly.
Looking for an intern, junior data scientist, or software developer? Let's connect.