Cloud Cards and OCI Control Center allows Customers or Users to track all major services historical usages across all regions. Organic Forecast notes these historical data points to predict what your consumption could look like in the coming months. This information is especially useful if the usage starts to breach the threshold and can allow you to make informed decisions for ordering more capacity. Service and Feature is available across 30+ regions @ OCI Customers.Designed, led and developed E2E solutions for Demand Planning and Organic Forecast solutions using AutoML features of Machine Learning leveraging Boosting, Bagging, ARIMA, Exponential Smoothing algorithms where without touchless Customers/Strategic teams can use Forecasting features and plan their Demands for their needs of Private or Public cloud services. Solution leverages OCI Cloud Native services and gets deployed via Terraform (IaaC) automatically whenever any new region or Customers get onboarded.Feature of the blog: https://blogs.oracle.com/cloud-infrastructure/post/organic-forecasting-demand-planning-oci-control-center

Project Name: Grievance Case Classification and Clustering Description: As a medical insurance organization, there are a lot of grievances that members send and a set to operation team reads and routes to appropriate queues manually. Solution saved $259k/year.Automated this process leveraging Azure OCR, Tesseract, NLP, information extraction and post-processing rule-based techniques to achieve an accuracy of >95% with an overall automation coverage of 75%. This solution is based on NLP, Deep Learning LSTMs and ML Supervised and Clustering techniques. Providing explainable features using SHAP and LIME frameworks.Project Name: PrePro AutomationDescription: As a medical insurance organization, there are a lot of claims and appeals that members send, and a set of Claims operation team reads the files and fills the data received in pdfs with image format manually. Entire solution saved $1Million/year for 2 queues. And projected to have savings of $2Million/year for all queues.Automated this entire process leveraging Azure OCR, Tesseract, Image recognition, NLP, information extraction and post-processing rule-based techniques. Leveraged a wide range of machine learning and deep learning techniques like CNN, NLP, majority voting classifier with a combination of Boosting and Bagging algorithms to extract 30+ fields automatically without human intervention. This automation enabled us to eliminate 50% of manual efforts with >98% accuracy.Project Name: Claims AppealDescription: There a lot of claims gets rejected and Members have privilege to raise an Appeal. An Appeal effort cost more than a Claim. And it would impact CMS (US Healthcare governance) ratings for the organization if Appeals turn out to be valid.Analyzed a billion samples with more than 300+ features and come up with a ML solution to predict if a raised Claim request can turn out to be an Appeal so that the Claims team can scrutinize more cautiously ahead. This effort eliminated 1000 Appeal cases/day.

• Identify existing Business and IT challenges in client applications. • Derive Opportunities from challenges and Developing use cases based on data analysis to improve business problems. • Transitioning of manual/repetitive tasks using AI/ML solutions. • Bringing value to Client Business and Applications. • Proposing and developing solutions using ML for Use cases derived.

• Eliciting Business needs and converting them into Business & Functional Design Document in the form of User story cards. • Co-ordinate with SME’s/ Client business team for further review and coverage. • Developing business requirements with Apex Classes, Triggers, and Lightning Components and Configuration changes on Salesforce. • Functional study and understanding of the applications and requirements. • Participation in meetings, Scrum, SME and Client business requirement meetings.