RAVI KUMAR NARUKULLA

• Well-qualified Data Scientist experienced working with vast data sets to break down information, gather relevant points and solve advanced business problems. Skilled in predictive modeling. Offering 12+ years including 7+ years of experience in data science and Analytics.
• Working in a product development framework and has good understanding on developing a data science solution with product mindset.
• Experience in solving data science problems in Agile methodology.
• Deep theoretical knowledge and experience in machine learning, deep learning, explainable AI and extensive experience in employing machine learning techniques for data driven information and decision systems.
• Excellent analytical, problem solving, communication and interpersonal skills with ability to interact with individuals and work as part of a team or independently.
• Good understanding on CI/CD pipeline and importance of ML Ops in data science world.
• Effective at communications with all levels of an organization, including technical, customers and management.
• Profound ability to handle end to end data science projects, big data analytics project for telecom domain.
• Experience in working with Green belt project using DMAIC Six Sigma methodology to reduce invoice reduction for EU region

Data Scientist and Big data Analyst (EMCDSA) from EMC2

Enhance Decisioning Product Capabilities:

As part of Decisioning Core team, my job is to research and develop data science components that enhances the product capabilities and delivers client projects in an efficient and faster way.

Worked on industry research and automated feature engineering for delivering faster propensity models using Web interactions data, Model interpretability and Automated machine learning, detector model and reusable feature engineering etc.

Customer Scoring for Marketing Campaign:

The objective is to score/rank customers that are high likely to make a purchase during the campaign. We have developed a scoring model which has given benefit of 150% higher response rate and 120% higher revenue than the average response in previous campaigns.

Customer Reactivation Model:

Objective of this project is to score deactivated customers that high probable to reactivate. We have developed a scoring model that can predict the probability of reactivation after 1 year of deactivation. We have achieved 2.5 time higher the customer reactivation after the model deployment.

ML for Digital transformation:

This project is project part of Digital transformation that we are making for the client who is into Chemicals Industry. Customer has a legacy system which keeps that information related to Billing, Transportation, Customer Service and operations data in a messages format. These messages (~1.5Million) contain the information of nearly 180+ attributes which needs to be extracted and loaded to the New salesforce system. We have used IBM Watson for Named extraction process and python for data standardization process.

Implementing Explainable AI :

The objective is to research and implement explainable AI capabilities to the existing ML pipeline. As part of this we have researched and experimented multiple explainable AI tools and techniques. Developed a standard framework for integration with existing ML pipeline and implemented LIME, Shap and global surrogate models.

Next best offer:

Deeloped a self-learning which can automatically recommend the next best offer based on historical acceptance. We have used multi arm based reinforcement learning which can automatically adjust offer (category) priorities based on the previous response and recommend next best offer.