Rajarshi Ghoshal

Help clients to determine price elasticity of CPG goods. Launched a module using Python, Pandas and NumPy to get user data and preprocess it for modeling step; used R for creating a linear model which will determine the relationship between price and volume of goods sold for CPG retail goods based on user input. Successfully ingested user provided data for modelling step based on metadata supplied by user. Created linear model in log space to determine price-volume elasticity and apply the model to generate price-volume curve. Used SciPy optimizer to optimize price point based on price elasticity coefficient.

Analyze data utilizing statistical and mathematical techniques. Develop machine learning models to automate and integrate processes. Formulate solutions to resolve problems in business processes. Write and apply algorithms. Collaborated on creation of machine learning system to boost process speed and increase results. Rated candidates according to relevance for job based on information gathered from experts on most valued attributes and skills for role. Applied expertise in Python, Panda, and PostgreSQL to developing solutions. Utilized NLTK grammar parser and LSTM-based skill extractor prototype to efficiently extract skills from job descriptions. Strategically executed named-entity recognition technique.

Assisted stakeholders with adopting machine learning practices and techniques to resolve business issues. Implemented machine learning and artificial intelligence to boost efficiency of existing solutions. Utilized Python, Pandas, Flask, Gunicorn, and SQL to create web application forecasting hiring costs. Formulated rule-based algorithm to predict various matrices in accordance with hiring manager and hiring location preferences. Proven success reducing uncertainty of hiring budget by ~30%, as well as manual labor within hiring process by 50%+. Designed indoor navigation system to assist with navigating through large retail stores, using Python, Pandas, Scikit-learn, and Flask. Ensured ability to identify location of individual based on Bluetooth Received Signal strength. Demonstrated ability creating and utilizing k-NN algorithm-based model via Flask to determine exact position of individual within store, with 74% real time accuracy. Obtained accurate location based on BLE RSSI of phone.

Developed knowledge of current data science practices, including Data Handling & Cleaning, Statistics, EDA, Predictive Analytics (Linear & Logistic Regression, Clustering, Support Vector Machine, Decision Tree, Random Forest), Time Series Analysis, Big Data, Model Selection, and Machine Learning. Applied theory in multiple projects, such as Healthcare Analytics project and Capstone Project. Acquired skills in various languages and tools, such as Python, R, and SQL.