Surya Teja Challapalli: Technical Lead-Analytics +91 6302045523
Surya Teja Challapalli: Technical Lead-Analytics +91 6302045523
Surya Teja Challapalli: Technical Lead-Analytics +91 6302045523
Summary
A Technocrat with almost 4+ years of experience executing data-driven solutions to increase efficiency, accuracy,
and utility of internal data processing. Experienced at creating Regression and Classification models,
Recommendation systems, Predictive data modelling, Image recognition, Object Detection/Computer Vision, and
analysing data mining algorithms to deliver insights and implement action-oriented solutions to complex business
problems.
SKILLS
Industry Specialization: Hypothesis Testing, Image Recognition, Object Detection, Market Basket Analysis, Item-
Item Collaborative Filtering approach, Telecom Churn Analytics, Credit Risk Management, Chat bot Integration,
Sentimental Analysis.
Machine Learning: Supervised, Unsupervised, Ensemble Techniques, Recommendation Systems, Deep Learning
using Keras/Tensor Flow, Computer Vision, NLP
Tools: Jupyter Notebook, IBM Data Science Work Bench RStudio, Zeppelin Notebook, IBM Watson Studio, Azure
ML Studio, Power BI, Informatica.
EXPERIENCE
Jan’2018 - Present: Technology Lead at Infosys Ltd, Hyderabad
Nov’2015 - Jan’2018: Project Associate at Cognizant Technology Solutions, Hyderabad
Feb’2015 – Oct’2015: Senior Software Engineer at HCL Technologies, Hyderabad
Oct’2010 – Jan’2015: Project Engineer at Wipro Ltd, Hyderabad
HIGHLIGHTS
Developed Market Basket Analysis based Recommender Systems for Blue Moon customer and increased
profitability in sales by 25%
Infused AI in Verizon telecom services. Reduced customer churn from 15-18%
Demonstrated POC on regression and classification experiments, and deploy models using Azure ML
studio.
Demonstrated POC on Chat bot using NLTK, verified and tested Infosys Nia application.
Published a detailed case study on marketing campaign data in Medium blog-
https://medium.com/@surya227239/artificial-intelligence-machine-learning-supervised-learning-and-
ensemble-techniques-6e6f25522ac5
CERTIFICATIONS
IBM Data Science Orientation Course
IBM Python for AI and Data Science
IBM Open Source Tools for Data Science
IBM Databases and SQL for Data Science
IBM Data Analysis using Python
IBM Data Visualization using Python
EDUCATION
Pursuing Post Graduation Program in AIML/Data Science, Applied Statistics from University of Texas, Austin
by 2020.
Bachelor of Technology in Computer Science at Andhra University, Vishakhapatnam.
PROJECTS
Market Basket Recommended Engine | Client: Blue Moon, Belgium at Infosys (Jun’19 - Present)
Outline Blue Moon is proudly brewed at the Sandlot, at the RiNo Brewery, and at breweries throughout
the world. For more than 20 years, the Sandlot brewery provided our team of brewers an
opportunity to create and share new beers with fans around the world from Denver. Rather going
for a beer, blue moon proposed a package in which food item can be covered with beer for
customers at lower price and want to increase its sales by running a recommendation engine.
Roles
Discussions with SME’s upon feature extraction from the historical bills.
Collaborate ETL team for converting raw XML bills to SQL tables.
Perform ML lifecycle (Data Pre-processing/Wrangling, Normalization, Visualization, Building model,
Evaluation, Cross Validation, Deployment).
Provide insights to customer.
Publishes the model into GitHub production.
Outline Verizon domestic billers (BODS) data was stored in SQL- server and VESDW data stored in Oracle
databases. Verizon faces a challenge of customer churn out to other telecom companies domestic
and global SP’s client wants to do Churn Analysis of customer with Proof of concepts based on
churn status
Roles
Worked on POC for Data Science and Visualization using Python (Jupyter Notebook)
Built experiment for Telecom Churn prediction using Azure ML Studio
Loaded scored datasets into Azure SQL
Built reports, applied visualizations on target column (box plot, pair plot, pie chart, table) formats
Provided insights to customer and published reports to Power BI cloud for easy access.
ICETS-NIA Chat Bot Framework Development| Client: Infosys Internal (Jan’18 - Dec’18)
Outline Infosys Nia provides a framework for enterprises to rapidly create and deploy conversational
interfaces in production. It understands natural language, responds accurately to user queries,
and is easy to deploy on a wide variety of messaging platforms, mobile apps, wearable devices,
smart speakers (such as Amazon Alexa, Google Home), and on web interfaces.
Roles
Worked with SME for keywords and wording patterns from Infosys database
Created chat bot application framework
Tested the chat bot framework and demonstrated NLP to business
Outline Horizon is reputed for its health insurance services and providers of coverage over 3.2 million
people throughout all of North, Central, and South Jersey. Fraud costs the insurance industry and
insurers are marshalling their data resources and creating a multi-channel approach to fraud
detection. Company is taking a very close look at claims and policy data. Entailed a predictive
model to crack down the suspicious claims.
Roles
Understand business specifications and apply EDA, statistical techniques
Visualized the model and provided insights for client.
Performed Data pre-processing, modelling, cross validation on binary classification algorithm.
Test the model in production and validate the performance measures.
Modified hyper parameters and tuned the model for better accuracy score.
Deployed the model into production environment.
Outline Horizon Blue Cross Blue Shield provides health insurance coverage to over 3.2 million people
throughout all of North, Central, and South Jersey and best known for their managed care and
traditional indemnity plans for individuals and employers. Informatica runs on ICOE, EDI and ILM
applications and was planned to migrate to Big Data due to vast of insurance data.
Roles
Worked on kernels, pandas, numpy in Jupyter Notebook
Tested the model in production and validated the performance measures.
Modified hyper parameters and tuned the model for better accuracy score.
HACKATHONS/CASE STUDIES
Deep Learning
o Built a Deep Neural Network model in Keras that classifies handwritten digits in MNIST dataset.
o Image Classification using CNN