Debojyoti Goswami

Helping a team of talented young data scientists prosper in their pursuit of data science model validation for a North American Banking Giant. In terms of hand-on- deliveries in the credit risk space, am responsible for:

• First Party Fraud Models
• Mortgage Reconciliation using NER in SPCAY
• Voice to text analytics using a large language model for banking call center to analyze customer sentiment and topics of concern
• Propensity and campaign management models for customer acquisition in the retail banking product spectrum
• Stress Testing Models
• Daily Balance Forecasting

Designed data science and machine learning solutions for a U.K. telecom giant that involved:

• Improving agent performance and overall and chat specific customer NPS in Chat Centre Operations using NLP in Python
• Sentiment and Intent classification of Chat Centre Data
• Determining topics of conversation between agents and customers from chat transcripts Market based pricing model for global access service
• Enterprise fraud detection in terms of finding out business customers that use consumer broadband for their operations
• Real time decisioning system enhancement by enriching online campaign management
• Next best offer models
• Fault SLA prediction
• Fault SLA failure prediction
• Impact of Weather Events on Broadband Quality of Service
• Customer survey analytics- sentiments and topics Network fault prediction
• Designed a Propensity to Call (PTC) model for customers that may call in for billing issues
• Churn propensity and churn driver analytics for Retail Consumer Broadband
• Order lead time prediction for Ethernet Orders
• MAC Address to vendor mapping
• Order SLA failure prediction

• Predicting the UNSPSC category of spend data and social media analytics for a petroleum giant in UK
• Healthcare analytics for a drug testing giant in US that involved analysis of KPIs in the efficient performance of clinical trials
• Benchmarking analysis of clinical trials for a drug testing giant in US, related to different clinical indications to identify areas of improvement
• Text mining of clinical trial protocols for a drug testing giant in US, to build a searchable database that will replace the use of reading clinical trial protocols with an efficient text search for the use doctors for reference, analysis of test cancellations

Analysis of hedge fund investor and fund manager data and design of inputs for a product to facilitate trading of hedge funds.

Data capture and analysis of hedge fund documents of Accredited Investors, Qualified Purchasers to map investor information into database.