Sajan Bhagat

Leading team of data-scientists and engineers to create HyperSense AI Studio(No code and low code machine learning platform). Developed DaskOrch an Orchestration Framework for distributed ML training using open source dask-kubernetes package. Build and productionalized an OCR solution for, digital automation of lease contract management system for Optus to replace legacy system. Improved efficiency of existing fraud management system by 8.2% by building Telecom Scam Detection Model to identify scammers in real-time. Reduced the handset fraud by 6.43% by building handset fraud model for telecom customer. Collaborated with Subex HR Team to develop an employee attrition model to improve employee retention. Created various eKYC solutions using computer vision for identity resolution and identity verification of digital customers. Leading data science community of practice.

Increased the customer base for vSphere Platinum by 5X using propensity to up-sell model. 30% customer upgraded to higher vSphere version as a result of propensity to upgrade vSphere model. Designed and built low touch and no touch system for deal management and deal approval. NLP and robotic process automation based cognitive bot to automate ticket resolution, honoured with “Best Digital Transformation Project” in Global OpEx2020.

An Impact of $144 MM was generated from List Price Optimization(LPO), by optimizing the price for 23K SKU’s. Developed Discount Guidance Tool to facilitate Deal Management Team to offer a range of discounts, for different combination of drivers of discounts obtained from Discount Driver Analysis. Achieved a Lift of 8x in top 2 percentile for Partner Propensity to Sell and captured not-transacting partner that attributed $7.68 MM per quarter on average. Built an in-house web crawler tool(Prospector) for VMware which reduced the quarterly spent made by VMware for external data.

Optimized marketing retention strategy to retain identified customers that can possibly iterate in future by predicting customer lifetime value (CLTV). Performed Survey Analytic and multi-regression models to understand drivers of NPS (Net Promoter Score). Built a product “IRX- Indian Realty Exchange” for client, later acquired by Quikr India for $60 MM. Designed architecture and developed Smart Deploy (utility for tableau workbook/reports deployment) from develop to production environment.