Mohit Raut

Experience

• 

  Technology Transfer Group, IIT Kharagpur

  Apr 2012 - Apr 2015

  • Student Co-ordinator

    Apr 2014 - Apr 2015

  • Student Head

    Apr 2013 - Apr 2014

  • Member

    Apr 2012 - Mar 2013
• 

  Rajendra Prasad Hall Association

  Aug 2012 - Apr 2013

  Technology Events Secretary
• 

  Notionink Design Labs

  May 2015 - Jul 2015

  Intern

  The aim of the internship is to design and simulate an electric motor.

• 

  ZS

  Jun 2016 - now

  ● Led the data science team for a B2B SaaS product enabled across 8 global clients incorporating customer personalized requirements ● Developed an AI/ML B2B SaaS platform aiming at predicting Opinion Leaders from concept to commercialization, bringing over $3MM revenue in 8 months and increasing Pharma Companies’ Key Opinion Leader (KOL) engagement efficiency by ~40%● Developed NLP based multi-label topic detection and novel entity sentiment detection modules using weak supervision and transformer models to efficiently capture social media perception of Opinion Leaders towards the client brands● Built a custom Physician network based on real world interactions captured across multiple open source and client datasets to identify local communities of highly connected physicians and local influencers using network analytics ● Managed a team of 5 data scientists and data engineers on a day-to-day level across multiple releases of the product● Developed ML-based Risk detection framework to help accelerate insights generation and designing actionable insights for Pharma Compliance showcased in Pharmaceutical Compliance Forum 2022● Leveraged semi supervised learning and isolation forest model to help accelerate salesforce risk detection and monitoring across the US market● Collaborated with a team of ML engineers to host and operationalize the solution on clients’ MLOps platform● Identified ~$18M (15%) incremental annual brand opportunity within US market based on predictive performance of a patient event prediction model designed to identify key drivers of patient attrition/non-adoption to curate interventions and increase patient drug adoption● Designed a time-indexed Physician Adoption model on highly imbalanced data (~2%) to identify potential physicians to target to increase conversion rate with 8x better recall than baseline predictions Show less Worked across US Life Science Firm projects and internal commissioned Proof of Concepts (POCs) with focus on solving business questions through novel ML and statistical based analytical techniques using Real-World Data Resource Optimization based on Dynamic Targeting:● Built and implemented patients’ RWD based suggestion engine to help identify “right customer at right time” for Field Reps ● Doubled the customer hit rate (% of alerts reaching right customers at right time) in Rep alerts● Leveraged time indexed modelling for patient event prediction using XGBoost and Bayesian Optimization to triangulate right customer to target ML Driven Anomaly Detection ● Designed and implemented an active learning framework on top of ML algorithms to identify trend level anomaly patterns via user feedback● Developed a Dash based interactive front end for user review of detected anomalies and deployed on AWS Sagemaker for biweekly refresh● Modelling solution based on Statistical Techniques (Chow Test, Bollinger Bands and Moving Averages), Unsupervised learning methods (Isolation Forest, PCA, KNN) and DL techniques (Autoencoders) Cohort Identification based on Survival Differentiation● Identified key drivers that increase/decrease the probability of experiencing long term benefit within patients in two comparator arms of Drug Therapy● Leveraged Propensity Score Matching and Multivariate Survival Trees on Real world data to identify the key subgroups with significant difference in survival ● Validated through gold standard clinical trials data Detection of Treatment Pathways in Oncology Disease Area● Designed an experimental framework to derive treatment patterns across Healthcare providers ● Leveraged ML techniques like PU learning, Clustering and rule extraction using Decision Trees to boost sample size for limited patient cohorts Show less Worked on solving sales force deployment, resourcing and customer segmentation problems for top 10 pharma companies in US through statistical modeling. Key projects● End to End Launch Analytics for leading Pharma client including Chart Audit, Sales Force Sizing, Customer Segmentation and Territory Alignment ● Customer segmentation for a breakthrough brand launching in multiple markets● Optimization of sales force touch points to effectively target the right customers● Go to Market strategy for brand launch in multiple EU markets comprising of 11 countries Show less

  • Data Science Manager

    Jan 2024 - now

  • Data Science Consultant

    Jul 2021 - Dec 2023

  • Data Science Associate Consultant

    Jan 2019 - Jun 2021

  • Decision Analytics Associate

    Jun 2016 - Dec 2018