Top skills - Website and mobile app development, data analysis, visualization of reports, building interactive dashboards
Technologies - CSS, HTML, Java, R, Python, JavaScript, Node.JS, React.JS, SQL
Strong expertise in machine learning and data analytics as well
I am familiar with Data Analysis aspects such as – Univariate & Bivariate Analysis, Missing Value Imputation, Feature Engineering, Categorical Variable Encoding, Normalization & Standardization, Binning, Outlier Handling, pre-processing of data & selection of right machine learning models.
• Develop data-driven insights & solutions with advanced analytics (machine learning, deep learning and statistical models) to extract insights from large proprietary data sources
• Identify, implement, and present results of research on ways machine learning approaches can aid decision-making
• Choosing appropriate Machine Learning/Data Mining algorithm
• Deploying Machine Learning models into production - Deploying machine learning model as Web Service: Building & training machine learning model, Storing the trained model in a pickle file, Creating REST API using Flask
• Deploying machine learning models to cloud - Building & training machine learning model, Storing the trained model in a pickle file (serialized format), Using Flask to build REST API based web-services, Writing Dockerfile to build the docker image
• Working on large datasets & coordinating with internal teams and clients through Agile methodology
• Experience in data visualization projects to build interactive business dashboards
KEY EXPERIENCES:
Client – Largest Investment Bank of UK
• Project 1. Modeled trading strategies using Pandas, Numpy, Matplotlib, SciPy, Scikit-Learn
• Responded to ad-hoc requests for insights from data with reports built using SAS, SPSS & Python
• Identified scenarios to predict trends in price movements. Calculated probabilities & expected gains following “Buy/Sell” signals generated by ML model (Random Forest, RNN & LSTM)
• Built a Random Forest with multiple Decision Trees to gauge the loan risk & to predict an interest rate
Client – Largest Retailer of UK
• Project 2. Did sales forecasting in Python with time series analysis tools such as ARIMA & SARIMA
• Used k-means clustering to group consumers of beauty products. Used Discriminant Analysis to spot demographic/behavioral traits to distinguish between users & non-users of products/services
• Built social media monitoring & analysis tools (NLTK) using NLP (Named Entity Recognition, Sentiment Analysis, Text Summarization, Aspect Mining & Topic Modeling)
• Used PCA to reduce dimensionality of large datasets
• Did Marketing Mix Modeling to analyze multi-channel promotional strategy. Created a random forest model with sales as the target variable and marketing channels as the feature inputs. Calculated the feature importance of each predictor and plotted it on a bar chart
• Used LightGBM regression to build the marketing mix model