• University of Colorado Boulder, Boulder, CO

• Manipal Institute of Technology, Manipal, India

I am primarily interested in:

• Machine Learning, Deep Learning
• Data Engineering and Data Warehousing.
• Data Mining and Statistical Analysis.
• Data Visualization.

• Languagues:

- Python3, R, Oracle SQL, Shell & Unix Scripting.

• Libraries:

PySpark, Keras, Numpy, Pandas, Matplotlib, Dplyr, Pytorch, Scikit-learn, Seaborn, Jupiter Notebook, OpenCV, Spacy, Data Mining Techniques, Machine Learning Techniques.

• ETL Tools:

Informatica BDM, Informatica PowerCenter, AWS Glue, AWS Lambda, Dataiku DSS, Autosys, ETL Pipelines, Data Warehousing, Big Data Pipelines, CI/CD process

• Developer Tools:

MATLAB, Git, VS Code, Microsoft Office, AWS EC2, AWS S3, Microsoft Azure, Tableau, APIs, DevOps, PowerBI.

• Big Data Technologies:

Hadoop: CDP, HDP, HDFS, Apache Spark, MapReduce.

• Project Development Methodology:

Agile and Waterfall.

September 2020 – July 2022

• Led automation of generating temporal business extracts; used ETL tools like Dataiku, Oracle SQL; reducing 50Hrs of manual work per extract and 75% of client expense.
• Spearheaded an interactive analytical dashboard to find a correlation between different databases from various data warehouses in python; reducing manual comparison by 25%.
• Developed ETL mappings to move data from multiple data sources into a data warehouse; simplified existing ETL mappings; decreasing run time by 15%.
• Led a team of 5 to develop Git pipelines for performing ETL using AWS Lambda, GLUE, Spark; cutting down manual work by 10hrs.
• Trained and mentored junior programmers in ETL methodologies; cutting down time for development by 20%.
• Directed various build activities aligning underlying architecture with business requirements, addressing business issues, data reloads, building data pipelines using Informatica BDM, AWS Glue, and Git pipelines.
• As an ETL developer, I am well-versed in Oracle SQL, Informatica BDM, Dataiku DSS, Pyspark and AWS services like Glue, Lambda, S3, and EC2.
• I also have experience operating very large data warehouses, Big Data technologies, designing schemas, relational & dimensional data modelling and working with slowing changing dimension tables.
• I have collaborated with product owners, and clients and have trained incoming freshers and laterals on our application and various other tech stacks. I also have exposure to visualization tools like Microsoft PowerBI and Tableau.
• I have built Data Pipelines with an ETL job running on both AWS Glue and Dataiku to perform ETL on business requirements. I also have worked as an UNIX admin and have UNIX and Shell scripting experience.

May 2019 – July 2019

• Developed a Generative Adversarial Network (GAN) in upscaling a low-resolution scale 10 satellite image to a higher resolution scale 30 image; enhanced image detection performance by 25%.
• Designed a Super Resolution GAN using ResNet with skip-connections and loss functions tweaked to optimize the model.

• 🎶 Spotify Music Recommendation system, University of Colorado Boulder

- Spearheaded a team of 4 to fabricate a probabilistic model for the Song data obtained from Spotify API. Developed data preprocessing pipeline for performing essential Exploratory Data Analysis (EDA). Performed Feature extraction to enchance features and to develop a song recommendation engine using K-Mean clustering in R. The model performed with an accuracy of 75%. Performed regression analysis to predict the popularity of the song by using multiple song features. Created a dashboard using PowerBI to further enchance the useability of the project.

• 📈 StockBuff – Stacked Temporal and Sentiment Analysis, University of Colorado Boulder

- Implemented Data Mining and market correlations to proactively predict stock market fluctuations using LSTM, mixed feature classifications and sentiment analysis. Gathered twitter data using Twitter API and financial data using Yahoo Finance API. Performed sentiment analysis on the twitter data using FinBERT model. Deployed LSTM for time series analysis, NLP features for Sentiment Analysis and used ensemble model to boost the model’s prediction performance in predicting a stock price of a company.

• 💲 Mutual Identity Authentication for Blockchain, Manipal Institute of Technology, India

- Developed an authentication protocol for mutually authenticating between three parties in a public blockchain system at any given time; reducing by 75% of authentication time compared to existing systems. Led the implementation of a five-node public blockchain network using Hyperledger Fabric and Golang; applied Elliptic Curve Cryptography for developing an authentication protocol.

• Received Spot award from Deloitte for leading a project, troubleshooting major issues and minimizing expenses for clients.
• Served as Executive Head of Electronics club, MIT, Manipal to organise different activities including general meetings, informative talks, and hackathons. Volunteered for Volunteering Service Organization (VSO) in uplifting the community.

• Super-Resolution of Level-17 Images Using Generative Adversarial Networks
• DDMIA: Distributed Dynamic Mutual Identity Authentication for Referrals in Blockchain-Based Health Care Networks

• GitHub: Nikhil-B-Madhu
• Linkedin: Nikhil B M
• Twitter: Nikhil B M