Data Scientist

The purpose of the bot is to help in understanding an automation solution for assisting the user to check the severity of Covid based on the symptoms the user has. It helps in assisting the user for taking the further steps based on the symptoms which is helpful in taking early decisions.

• Developed in Dialogflow CX
• Dialogflow CX described and visualized as a state machine, configured to collect information or parameters from the end-user
•  Our bot enabled our end users to come back to the chat anytime and continue from where they left off
•  create a custom implementation using a webhook wherein a function will store the parameter and forms you collected and use that to continue the chat from where the user left off during a conversation flow or a session.
•  In the webhookResponse you can set the fulfillment_response,target_page fields and session_infofield to update and send back the stored parameters you collected from the previous conversation
•  Conducted complete analysis of system and business requirements for all IVR applications
•  Resolve complex technical design issues by analysing the logs, debugging code and identifying technical issues/challenges/bugs in the process

The purpose of the bot is to automate the customer service desk activities by answering day to day activities and guiding them to resolve the issues by providing solutions in detail like payment details, Billing etc.

• Developed textual and IVA based models using Dailogflow CX.
•  Built the complex bot with many flows, pages, parametric conditions, cloud functions, webhooks etc, used all the functionalities in Dialogflow CX to build this complex architecture or flows.
•  This Bot accelerated customer support for similar queries via automation.
•  This complex bot helped in reserving the human resources for even more complex, productive and creative tasks.
•  Enhanced the overall efficiency by minimising the delay in transferring tasks and achieved the optimized performance with less effort in less time.
•  This complex IVA based bot with the help of dynamic functionality feature in Dialogflow CX progressively escalated the business higher with lesser hassles.
•  Minimized the requirement of customer care resources while improving quality.

The purpose of the bot is to automate the customer service desk activities by answering day to day activities and guiding them to resolve the issues by providing solutions in detail.

• Developed in Dialogflow CX
•  Developed a chatbot that improved the scaling operations of the organization like supporting many numbers of customers at same time without any abruption
•  Developed chatbot with advanced features like IVA or Voice Bot using the advanced agent service like Dialogflow CX which is very suitable for large and complex agents. It includes flows, pages are the building blocks of the conversation design. It uses the dynamic nature of NLP appropriately
•  Optimized the developed chatbot by storing the data while interacting with the customers and analysed which helped in improving the customer experience by making use of dynamic functionalities of Dialogflow CX
•  Developed voice-enabled Bot which uses spoken dialogue from users as input that prompt responses or creative tasks
•  The developed chatbots played a valuable role in various vital processes, especially by driving the customer engagement with 80-90% response rates. This helped the customer support team to
  concentrate more on complex customer issues
•  This bot simulated an intelligent conversation with multiple users at a time via IVA and textual methodologies foo engaging in conversation

The purpose of the bot is to update the user information like phone number, job title, company name, location, pin code by sending a link to their mail to maintain info.

• Installed and configured Apache HTTP Server
•  Installed and configured the Stanford NLP Server
•  Configured SSL on Apache HTTP Server
•  Deployed the application on AWS as webapi.
•  Implemented CNN, RNN, SVM algorithms.
•  Collaborate with data architects, Team members on project goal
•  Actively Involved all the in-design discussions to handle various responses from users.
•  Developed a Flask API to handle all these questions

• Main objective of the Project is to predict the demand of the product at various levels and predict whether a customer will come in next floor set or not, if so when they will come back and what they are more likely going to purchase
•  Design an approach to predict whether a customer will come back or not
•  Involved in design discussions for various functionalities like Data transformations, Data cleaning and deriving new feature
•  Built high performance models and performing hyper parameter tuning.

Main objective is to predict the claim rate, Renewal Rate for week over week.

• Designed an approach to predict whether a customer will renew or not
•  Involved in design discussions for various functionalities like data transformation, data cleaning and deriving new features.
•  Setup R studio, Jupyter notebook on AWS
•  Involved in designed discussions to handle the alerts and reasoning questions for week over week
•  Implemented XGboost, Random forest, ANN
•  Collaborated with Data Architects and team members to achieve the project goal.
•  Built high performance models and hyper parameter tuning using GDA, MGDA

• Implemented for mobile application
•  Implemented Apriori, algorithm, FP-growth, K-means, Sentiment Analysis.
•  Collaborated with Data Architects and team members to achieve the project goal.
•  Built high performance models and hyper parameter tuning using Adaptive learning and genetic methods
•  Used different methods to handle, outliers, missing values, duplicate values.

• Algorithms used
  • Decision trees
  • Radom Forest
• Logistic Regression.
  Main objective of the Project is to predict whether the Customer is getting Delinquent and Default or not and its delinquency rate which helps the bank incurring loss
• Business understanding and Data Ingestion
•  Feature Engineering like output column derivations using the existing features, feature cleaning, feature learning
•  Used algorithms like Random forest, decision trees and KNIME for feature selection
•  Handled the project individually on achieving the project goal
•  Built high performance models and performed hyperparameter tuning using adaptive momentum - based algorithms

• Algorithms used
  • ANN Radom Forest
  • Logistic Regression
• Understanding how Banking domain works, business understanding and Data Ingestion.
•  HandledhighlyImbalanceddataproblemwithownapproachandcomparedwithother techniques like SMOTE
•  Performed Exploratory DataAnalysis forbuilding efficientmodelsand drawing the insights Handled the project individually on achieving the project goal
•  Used advanced algorithms likeGenetic algorithms for feature selectionwhich resulted ingreat accuracy
•  Built high performance models and performed hyperparameter tuning using algorithms like Gradient decent algorithm and momentum based gradient decent algorithm