5 years experience in Data science using Python, R, SAS, Julia.
- Master of Science, Mathematics, Indian Institute of Technology, Delhi (6.725/10)
- Bachelor of Science, Mathematics, Govt. College, Rourkela (79.4%)
- 12th , Kendriya Vidyalaya, Bondamunda (76.6%)
- 10th , Kendriya Vidyalaya, Bondamunda (74.8%)
Lead Data Scientist
PropheSee Pvt Ltd.
(Jan, 2016- present)
Leading all Data Science Projects at PropheSee, which allows brands to leverage their digital presence by developing actionable insights, predicting outcomes & optimizing for success.
- Prepared a classification algorithm which is ensemble of 3 distinct machine learning algorithm (KNN, SVM and Random Forest) which classifies all social media post as paid or organic.
- Used publicly available data to predict private level metrics on social networks, this is an estimation problem on which
- Designed a model which scores social influencers in the social network, used Machine learning algorithms to create such algorithm.
- Developed Web hooks model which has ability to predict the total engagement any post will, Build an API to implement the algorithm on website. This is based on Machine learning and Time series algorithm.
- Build a grading system which benchmarks your performance against your industry. It further recommends the key insights to be looked into in order to improve.
- Prepared sentiment analysis using NLTK, which provides sentiment for each costumer post on your website, also prepared Verb and Noun extraction models using Natural Language Processing which helps to extracts such information for any band
- Tools used: Python, AWS, Unix
Sr. Statistical analyst
RSG Media System Pvt Ltd.
(Apr, 2014- Dec, 2015)
Working with a US based media client. Responsible for creating Statistical and Mathematical models using Python, Julia, R, Mosel and multiple open source add-ons:
- Developed an Operation research based model using Julia which schedules promotions (for episodes) optimally and eventually increases the viewership. Currently this model has given 10% lift in the viewership for the targeted Demography
- Cross channel campaign management: Scheduled promotions to expose them across multiple network which significantly increased the conversion of viewers.
- Creating various Machine Learning models using R to maximize the unique number of viewers watching a program.
- Used Julia to build a simulation model which calculates the average number of unique viewers watching in a particular time frame.
- Developed gradient descent method based logistic regression model using R which can predict the customers who are likely to watch the program after watching the promo of the campaign.
- Working on a time series model which can predict TV ratings for a given set of Programs
Absolutdata Research & Analytics
(Jun, 2013- Apr, 2014)
Working with a US based hospitality giant having SQL based data spread over more than 100 different dataset (tables). Responsible for creating various statistical models using SAS to target a fruitful group of customers, Designing optimized emailing vehicles containing an optimal number of offers for a customer. Build multiple models which predicts the propensity of a costumer to book a property
- Created a Machine learning algorithm based Propensity model which predicts the geography in which customer is most likely to stay in a particular month using SAS and SQL.
- Created an Emailing channel which targets the customers who are attracted to last minute offers.
- Developed a VBA based interactive simulator to optimize the Travel Destination special emails.
- Awarded the Star of the Month award for being the best performer in October 2013
Genpact (Sep, 2011- Jun 2013)
Responsible for analyzing the needs of the business, customers and coming up with solutions to business problems. Involved in the managing multiple small projects, using SAS, Excel, VBA, and UNIX to enhance the existing functionality for global client. Managing the work from initial request to estimation, approval, and analysis using SAS and Excel.