Principal Data Analyst
Resume:
Madhur Devkota
Permanent Resident, authorized to work in the US without sponsorship
Tableau Dashboard
www.madhurdev.com
******.***@*****.***
github.com/madhurdevkota
linkedin.com/in/madhurdev
PROFESSIONAL SUMMARY
Principal Data Analyst with a decade of proficiency in Python, Machine Learning, and advanced Geospatial analytics. Specializing in transformative projects that enhance organizational efficiencies and advance ML adoption across sectors, with a strong foundation in geospatial technologies and data science that enables groundbreaking insights and strategic innovations. Actively involved in research projects, has presented at conferences, highlighting the real-world benefits of Geospatial analytics; dedicated to data-driven research and continuous learning.
SKILLSET
Data Science: Advanced proficiency in Python, statistical analysis, machine learning, and deep learning libraries, Python Geospatial libraries. Scikit-Learn, OpenCV, TensorFlow, Keras, GeoPandas, Shapely, Rasterio, ArcPy, PyQGIS, Jupyter Notebook, Dataiku, DeepNote, MLflow, Postgres, PostGIS. Data Analytics & Visualization: Expertise in Tableau, Power BI, Python viz. library - Matplotlib, Plotly (Dash), Seaborn, ggplot2.
Geospatial Technology: GIS, Remote sensing, ArcGIS Pro, QGIS, ERDAS Imagine, Python, R, GeoPandas, Shapely, Rasterio, ArcPy, PyQGIS, GDAL
Civil Engineering: AutoCAD, Project cost estimation, structural drawing, hydraulic modeling, Project management. EMPLOYMENT HISTORY
FINRA Principal Data Analyst
Sep 2022 - Present
• Led data-driven decision-making by translating complex business challenges into scalable analytics solutions, enabling leadership to make informed strategic choices
• Developed SQL-based data pipelines, optimizing queries for large-scale financial datasets to improve processing efficiency and insight generation
• Designed and maintained interactive dashboards (Power BI, Dataiku and Tableau), transforming raw data into actionable insights for stakeholders across audit, risk, and compliance teams
• Defined, built, and analyzed key performance indicators (KPIs) to assess financial risk, fraud detection effectiveness, and operational efficiency
• Conducted data wrangling and transformation using SQL and Python, ensuring clean, structured datasets for advanced analytics and machine learning models
• Presented insights and reports to senior leadership, distilling complex analyses into clear, data-driven narratives that influenced business strategy
• Designed and implemented data governance frameworks, ensuring accuracy, consistency, and compliance with financial regulations
• Optimized ML models, integrating geospatial and time-series analysis, improving fraud detection accuracy by 10-15%
• Managed ML pipelines on the Dataiku platform utilizing Postgres, PySpark for big dataset
• Developed novel ML/DL algorithms for detecting financial fraud patterns, improving detection by 30-35%
• Developed NLP solutions for topic modeling algorithms (BARTopic using HDBSCAN algorithm) to classify invoices based on content, enabling automated categorization and streamlined processing
• Conducted A/B testing and statistical analysis to measure the impact of business initiatives, translating results into concrete, actionable recommendations
Norfolk Southern Corp. Sr. Geospatial Data Analyst Atlanta, GA Apr 2019 - Aug 2022
• Led multiple Data Science projects, collaborating with stakeholders, and developing Data Analytics Dashboards in ArcGIS Dashboards and Tableau
• Designed and implemented geospatial data models and schemas to efficiently store, organize, and retrieve information, developed Linear Referencing tools
• Conducted data analysis using supervised/unsupervised techniques (Logistic Regression, SVM, KNN, clustering)
• Conducted comprehensive geospatial data analysis and developed advanced machine learning and time-series models to optimize freight transportation, improve operational efficiency, and forecast demand for better resource allocation
• Investigated land cover change using satellite imagery and elevation data, implemented Deep Neural Network architectures for image classification
• Worked with ArcGIS Enterprise, Oracle SQL, geodatabases, Postgres
• Developed Python scripts, geoprocessing tools, ArcPy scripts for workflow automation Mississippi State University Research & Teaching Assistant MS, USA Jan 2016 - May 2018
• Developed Geographically Weighted Regression models for Ocean Acidification using satellite data, raster product calculations, and supervised classification with SeaDAS and Python
• Employed advanced spatial statistical techniques like Regression, Geographically Weighted Regression, and Spatial Clustering using Python and GIS applications
• Built predictive models for ocean parameters using Hierarchical data formats (HDF, NetCDF)
• Enhanced student learning in GIS data creation, management, and spatial analysis, provided detailed feedback on assignments, and supported the academic advancement of up to 30 students per lab through personalized mentorship ISTEM Lab. Pvt. Ltd. Geospatial Data Analyst Kathmandu, Nepal Apr 2014 - Dec 2015
• Developed ML models for spatial processes using Python
• Performed geostatistical analysis on multiband raster data using Spatial Analyst tools
• Worked with rasterio, earthpy for climate modeling, multispectral remote sensing data
• Piloted Disaster Management System with mobile data collection, cleaning, spatial analysis
• Generated spatial analytical reports, schematic maps, advanced cartographic maps Swiss Agency for Development and Cooperation Civil Engineer Nepal Sep 2013 - Mar 2014
• Led Structural design and drawings, finalizing design reports by drafting AutoCAD drawings, and preparing rate analysis and quantity estimations
• Performed innovative Geospatial Spatial Analysis in Water Resources Engineering
• Conducted Geospatial Water resources analysis to estimate the Catchment area for hydrological investigation at bridge sites
RESEARCH
• Adopting Machine Learning Unsupervised classification techniques for Spatial Water Quality data (Research completed, Current in writing) - GitHub Project URL
• A geographic weighted regression approach for improved total alkalinity estimates in the Northern Gulf of Mexico - Journal Publication Url
EDUCATION
DataCamp Machine Learning Scientist using Python Cornell University, NY, US Satellite Remote Sensing using Python Mississippi State University, MS, US Master of Science in Geospatial Data Science National Institute of Technology, India Bachelor in Civil Engineering
Contact this candidate