Rutvik Parmar

• As a part of Amazon's Social Media Capacity Planning Team responsibility was to develop and maintain data mart to support BI requirements and enable business decision making by creating reports in Quicksight/Tableau for both Demand (Customer Contacts) & Supply Side (Customer Service Agents).
• Developed Daily/Weekly/Monthly business reports focusing on contacts actuals, forecast, its variances and service levels. Also provided stakeholders with deep dive views for CS Agents performance metrics.
• Developed Python Modules on AWS to generate baseline forecast commentary to support forecasting analyst to provide outliers for the business.
• Worked on building near real time data patterns for ingesting Social Media Contacts (Extracting Customer Queries from Platforms like Twitter, Facebook, Instagram, etc) from Sprinklr Systems. This involved using AWS Lambda, S3, Redshift.

For Chubb Limited

• Understanding key paradigms of Insurance Domain and propose, build and maintain system design and flow to handle large historical claims data stored in legacy systems.
• Write SQL Queries, Stored Procedures and SSIS Packages to ETL the data from legacy systems to data modelling.
• Develop star schema model, storing fact and dimension tables to be used by Powerbi to create claims dashboards.
  

For Rivittv (December 2019 - December 2020)

• Using Lambda Architecture to process both transactional data and realtime event data using Spark Streaming.
• Developed Data Pipelines in AWS to load data from AWS Aurora Database and Store it in Amazon Redshift.
• Finally performed data warehousing to create fact and dimension tables to be used by Tableau to create dashboards.

For Infinis Energy (September 2019 - December 2019)

• Used incoming SCADA data to create reports using PowerBI which enabled near-realtime monitoring of gas engines and developed ML models to predict time to failure.
• Performed Feature Engineering and Implemented PCA algorithms for dimensionality reduction and feature selection.
• Used XGBoost and ARIMA models deployed in Azure and reinforcement learning to continuously improve model performance.
• Overall reduced Infinis Energy operational cost and lowered Schedule maintenance.

For Ferrero India Pvt Ltd (April 2019 - September 2019)

• Developing custom Power BI dashboards for global Ferrero brands.
  KPIs in dashboard helped Ferrero’s stakeholders to identify their sales pattern across different product category.
• End-to-End data driven solution was developed wherein we used SSIS for ETL, SSAS for data models and Power BI as front end reporting tool.

For Suzlon India Pvt Ltd (July 2018 - March 2019)

• Introduction to Wind Turbine Components such as Pitch Bearing, GearBox, Generator and Main Bearing.
• Implemented Azure Storage Components such as Blobs, Queues, FileStorage and Azure VMs.
• Built business solutions to deploy custom R Scripts onto Azure ML as a web service.
• Debug custom R Scripts and optimise it to work with Azure ML Environment.

For Walgreens Boots Alliance (February 2018 - June 2018)

• Introduced In-Memory computation including Apache Spark and in-memory databases to the organisation. Socialised use cases and build consensus for adoption of new technologies and open source.
• Built business case to do social media analytics and sentiment analysis using NLTK.

For MuleSoft (September 2017 - December 2017)

• Read API log data from AWS S3 buckets
• Evaluation of API data using log data to predict anomalies in API usage. Used Support Vector Regression and SARI-
  MAX and RNN to evaluate models with the best results.
• Used plotly for visualisation of anomalies. Used Anaconda for code development.
• Used KMeans to detect anomalies based on IP address.
• Exposed data pipeline as REST API to convert prototype into production code.

For CVSHealth (August 2016 - August 2017)

• Used Kmeans for identify clusters within data for pattern analysis. Used Weka and mahout for supervised learning. Used Tableau for data visualisation.
• Conducted training sessions on Apache Spark, Machine Learning algorithms such as SVD, Collaborative Filtering, Kmeans clustering, Predictive Analytics using decision trees such as Random Forests, Cluster analysis and ANOVA, and classification using Support Vector Machines (SVM). Provided guidance on when to use what algorithm.
• Design web analytics solution using Lambda architecture based on Hadoop to analyse click stream and web logs using Apache Spark, Kafka, Pig. Implemented data pipeline operations using Pig and used Kafka for publishing click streams on HDFS. Used Tableau for visualisation.
• Design social web mining solution by subscribing to streaming Facebook and twitter feeds and Natural Language Pro- cessing (NLP). Customised NLTK corpus to bias with NDC drug codes to increase efficiency.