Download as pdf or txt
Download as pdf or txt
You are on page 1of 3

Aditya Upadhyaya Email: adityau007@gmail.

com
Business Analyst/Machine learning/Analytics LinkedIn: https://in.linkedin.com/in/aditya-upadhyaya
Gurgaon, Haryana – 122001 Contact: +91-8754117823/8939774321

Summary

• Expert in developing Insightful and Business focused Statistical Data Modelling, Predictive Modelling, Data
Analytics solutions using Machine Learning, SAS, Python, R, Hive, PySpark.
• Excellent knowledge of EDA, Statistics, Machine Learning, NLP, Python, ETL, BI and related Data
warehousing concepts.
• Expertise in using databases SQL server, PostgreSQL, Hadoop to process data from source to reporting
layer for developing Analytics and Mining solutions. Clear understanding of all SDLC phases.
• Basic Knowledge of Deep learning, Big Data, Hadoop Concepts, Map-reduce and HDFS, Tableau, AWS SageMaker,
S3.
• Working experience in Risk Analytics, Banking and Financial Services, Life Insurance and Investment,
Travel and Entertainment, Research, Requirement Analysis, Application development and Maintenance.

Work Experience: 5.8 Years

Evalueserve Pvt Ltd.


Senior Business Analyst (Current Employer – Since July 2019)

EXL Services
Senior Business Analyst (July 2017 – June 2019)

Cognizant Technology Solutions


Programmer Analyst (April 2014 - June 2017)

Technical Skills:

Statistical Data Modelling, Exploratory Data Analysis, Business Analytics, Machine Learning,
Technology
Predictive modelling, Descriptive/Inferential Analytics.
Languages Python, SAS, R Programming language, SQL, Hive Query Language, PySpark, UNIX
Databases SQL Server, PostgreSQL, Hadoop, Teradata

Tools SAS, Python, R Studio, Octave, MATLAB, Pentaho (5.2), Tableau (8.x, 9.x), Informatica

Certificates and Awards

➢ Stanford University (Coursera): Machine Learning


Machine learning practitioner, March 2017
➢ John Hopkins University (Coursera): R programming
R programming practitioner, July 2016

Innovation Award: Awarded in Q1 2019 EXL Analytics R&R Awards for breakthrough in automated insights
generation using state of art analytics techniques.
Mastermind Award: Awarded in Q4 2017 EXL Analytics R&R Awards for quick adoption and
contribution to ongoing work.
Marathon Learner: Awarded in Q3 2016 GTO Global Awards for outstanding 102.60 hours of learning in a quarter
from cognizant academy web portal.
Innovation and POCs:
• Automated DOCx and PPTx and EXCEL processing for Named Entity Recognition, Content Identification,
Automated PPT generation. (Media/Banking Client) NLP and String-Matching capabilities, 60%-100% reduction in
Manual efforts.
• Data Enrichment using Python Web Crawler and Google Map API. (Banking Client) Reduction in Manual effort by
80%.
• Executive profiling using audio files from call centers interactions. Google API, NLP and Python has been used for
extraction, translation and feature extraction.
• Joint Value Proposition (Banking Client). A heavy data centric module focused of corporate travel expense.

Projects:

Client: Life Insurance Services (Fortune 100)


Name: Next Best Offer
Tools and Tech: Hybrid Recommendation Engines, Vector Similarities, Python, SAS, SQL
Business Problem: Development of methodology to propose next best offers (policies/riders) to current
clients.
Description: An exhaustive module combining Recommendation engines, cosine similarity and business
requirement to develop an application that can aid the sales team and executive while suggesting a next best
offer to existing clients.

Client: Banking and Financial Services (Fortune 100)


Name: Card Member Health Index
Tools and Tech: Unsupervised Anomaly Detection, Python, Hive, PySpark
Business Problem: Development of methodology to score and capture non-compliant CMs.
Description: Project is aimed at identifying, quantifying and targeting highly noncompliant corporate card
members so as to reduce reputational risk and associated monetary losses. Given no historical score/target was
present, PCA followed by similarity analysis was used on transaction level data. After profiling of produced
scores, CART was used to find underlying association between transaction level flag and non-compliance
behavior of card members.
Name: Automated Expensing System
Tools and Tech: Image Classification, CNN, Deep Learning, Optical Character Recognition (OCR), Python
Business Problem: Development of a completely automated expensing system.
Description: Project has two main components, Identification and correction of defects in receipt images and
Optical character recognition, validation and processing of final corrected images. CNN LetNet architecture has
been used for identification, classification of defects and Python tesseract is used for OCR. This is aimed at removal
of complete manual effort with integrated securities and behavioral fraud capturing capabilities.
Name: Smart Insight Generation
Tools and Tech: Natural Language Processing, Python, SAS, Statistical Data Analysis, Hive, PySpark
Business Problem: Smart Insight generation with no human intervention. Description: As an enhancement
to available T&E solutions for global corporate clients, this project is aimed at providing recommendations and
smart insights along with dashboard. Explanatory data analysis is performed on SAS to produce several
summary files which are used as Input for Cost driver analysis and Root cause analysis to find various Causes
and their Impacts highlighting the pain points and actionable areas to increase savings and Overall strength of
relationships.
Name: Expected Payment Risk
Tools and Tech: GBM, Ensemble Modelling, Predictive Modelling, Model Validation, Python
Business Problem: Predict the probability of a CM being delinquent.
Description: Module was aimed at identifying card members with high probability of going into next stage of
delinquency and eventually to credit loss. Leveraging the historical data of spend and payment of card holders, a
classification model was built to assign a probability score. Challenges involved were Data restrictions, Skewed
and fewer events. Terminologies used are Oversampling/Under sampling, Principal component analysis,
Clustering, CART, Logistics regression, GBM, KS, Gain and Lift, Model Validation Techniques.
Name: Merchant Name Matching and Sales Effort Validation
Tools and Tech: NLP, Python, Similarity Scores, Text Classification, Monge Elkan, Hive.
Business Problem: Develop a universal Name Matching algorithm to support strong
campaign distribution and trust. Classify the efforts entered by sales representative in form
of text as valid/invalid effort.
Description: Pertaining to telemarketing, Direct mail and Email B2B campaigns, Client’s US team needed to cross
check and verify each and every prospect credential received by various sources so as to decrease the discrepancy
in delivery of campaigns. Various NLP and string- matching algorithms were tested and Monge Elkan was
implemented with a specific algorithm for data processing. Extensive Functional Testing and documentation was
involved to make sure no stone is left untouched.

Client: GTO Research and Development


Name: Market Basket Analysis for association of skills
Tools and Tech: R programming, SQL, Apriori algorithm for Association Analysis
Business Problem: Associate Skill sets with one another from various job requirements to create clusters.
Description: cluster of skills generated by performing basket analysis of job requirements. The clusters generated
were stored in SQL database for future associate mapping requirements. The outcome of development task was
single R script file capable of interacting with SQL DB for inputs and output along with performing all the
necessary checks and analysis.
Name: Adaptive tool for Skin Disease Identification – App de Pele
Tools and Tech: Image recognition, CNN, MATLAB, Python, OpenCV
Business Problem: Development of Android App for identification of various Skin Disease. Description:
Aimed at full scale development as a health care product enabling users to check and verify the quality of skin and
several skin related disease with help of image captured through the camera of cell phone, this project involved
extensive research and effort for collection of data, Data processing, ROI extraction, Feature engineering,
Convolutional Neural Network.

Client: Cognizant Gain Analytics Core Team


Tools and Tech: ETL, BI, Development, QA, Production Deployment, DB2, PostgreSQL, Putty, HTML, Eclipse,
Fusion Charts, Report Designer.
Business Problem: Development of dashboards corresponding to various KPIs.
Description: Current performance, cost and associated metrics such as PPM (People per Meter), RFR (Revenue for
Resources), EPC (Expected Payment and Cost) etc. need to be monitored by CXOs of any organization. My
involvement for initial 2 years of career has been with preparation of ETL jobs, Reports, Integration of reports with
charts and HTML properties, publication of reports, development, testing and maintenance of developed
dashboards.

Academic Qualifications:
Degree/Education Institute/College/School University/Board Passing Year Marks Obtained
MSc. Business Pursuing - 4th
BITS PILANI BITS PILANI 2020
Analytics Sem

B.Tech (Electronics & ABES Institute of Uttar Pradesh


2013 64%
Communication) Technology, U.P. Technical University

Anil Saraswati Vidya Mandir,


Higher Secondary CBSE, Delhi 2008 73%
U.P.

Anil Saraswati Vidya Mandir,


Secondary CBSE, Delhi 2006 83%
U.P.

Strength & Weakness:


Confident, Punctual, Optimistic.
Emotions, Selflessness.

Hobbies:
Cooking, Writing, Poetry.

Date of Birth: 16 January 1992

Declaration: I hereby certify that all the information provided above is true to the best of my knowledge.
Signature: Aditya Upadhyaya

You might also like