Predictive Modellingo HCP behavior▪ Designed and implemented machine learning systems for Fortune 500 clients to predict prescribing behavior of health care professionals (HCPs) for new product launches.▪ Worked with tech team to implement a system which refreshed these models periodically and generated alerts to the sales reps whenever a new HCP in a geography is predicted to prescribe the drug.o Patient opportunity▪ Designed a machine learning model to identify patients likely to get on a therapy in future. This model helped in predicting patients who will switch to the therapy, and the new patientswho might try the therapy.o HCP sales▪ Designed LSTM based regression models to predict future drug sales at HCP level.▪ The results of these models were used to identify potential high value HCPs to be targeted by the sales force.Targeting and call planningo Identified relevant targets (HCPs) for specific drugs using patient level data (claims data) and sales data.o In case HCP – account affiliations were unavailable, used geocoding to cluster HCPs together which were accurate enough to pinpoint the actual accounts to be targeted.o Created call plan for sales force to target the HCPs identified above.

Customer Insights (Market Research)o Designed and successfully conducted qualitative and quantitative market research studies in multiple countries across US, CA and EU, to track awareness and current/future usage of pipeline and in-market therapies.