Sr. Data Engineer
Location : Jaipur, India,
Experience : 5 Year
Rate: $20 / Hourly
Availability : 1 Week
Work From : Any
Category : Information Technology & Services
Project Name: Dentine Informatics
Role: GCP Data Scientist | Dental Product
As a fraction of the Analytics team, I help dentine informatics dept. make informed decisions with contemporary AI & MLOps solutions leveraging GCP.
Projects: Performed Dental X-ray image classification with Fast-RNN, Transfer learning & ResNet to identlify decay types. And deploy it on GCP.
Project Name: CIVICA
Role: Azure Data Scientist | Housing Product
Campaign Strategy initiatives for internal products: Developed a Campaign Response Model to determine which customers should we target.
Developed Behavioural Segmentation (based on RFM data on how the customers buy) & Attitudinal Segmentation (based on what they buy).
Developed product class level Collaborative filtering for next best product as a part of Campaign Design Strategy for the target segments. Products in Quantitative Finance - US Post Trade financial client:
Forecasted Business Volumes (Deri-Serv, Loan-Serv etc.) & Server Capacity Planning.
Predicted Cross Sell propensity for mortgage loan customers in buying payment protection (insurance) products for US Banking client.
Developed predictive model for how likely a customer to buy a different payment protection (Insurance) Product using Multinomial Log-Reg.
Predict the likeliness among the payment protections namely - Single, Group, LCI.
Predicted Sales for a product category and sub category of CPG Client for modern trade and traditional trade markets.
Developed ROMI based model for TVCs using Multiple Linear Regression.
Products in Public sector – Health Care & Housing:
Rent Arrears & Chatbots - Created a module of Arrears Analytics to extract signals that could help managing the rent accounts and arrears efficiently.
Predictors for Pensions Department: Identified the Predictors/Signals of Retirement for the Pensions Department for Civica UK Finance team.
Client - Anheuser Busch InBev: To Assist different Zonal FP&A team forecast the beer sales in
the next 3 years to help the Business Units estimate company performance & incorporate
growth strategies in their annual budgeting process.
ML Model Used: ETS, Neural Nets, ARIMA, SARIMA, FARIMA, TBATS, Holt-winters, NAÏVE
BAYES, LOG-LOG, PROPHET, K-Means Clustering, LSTM.
Forecasted Beer Volume: Estimated the influence of seasonal flags from Macroeconomic, Quarterly, Covid, Lockdown, Expenditures & Weather data.
Forecasting P&Ls: Calculated the Marginal Contribution forecast for 2020-21 via
univariate to estimate GTO, Revenue, Discount, Logistics, Excise duty.
Forecasting Recent trends: Devised a daily-trend algorithm for 7day, 15 day etc. training
periods to improve & estimate the recovery of volume sales.
COVID analysis: Predicted the spread of COVID in Latin American countries with SIR
curves to provide an estimate of the economy’s recovery.
Quantified the dip in sales by super imposing the macro-economic assumptions to forecast the volume.
Geo-Spatial analytics:
High end segmentation: Estimated - 20K sales targets for selling High end products across Mexico City with clustering and dendrograms .
Discovery of new Sales prospects: Identified - 80K potential locations for targeting nonalcoholic beverages in a country.
Devised an algorithm to rank the locations based on its popularity in social media & did customer segmentation/profiling with K-means.