SOUMYADEEP NAG

Dynamic professional with experience at Amazon as a Catalogue Associate Analyst, where I enhanced AI accuracy by 95% through effective training and analysis. Skilled in quantitative analysis and project management, I excel in resolving customer disputes with empathy, driving satisfaction and loyalty. Proven ability to implement process improvements and maintain compliance.

PROJECTS

SIM Fraud Detection | Detecting Fraud using AIML

• Managed detection of masqueraded products to ensure alignment with Amazon product policies on a DART project. Implemented AL tools for further detection and classification of non-compliant products.

Redrive | Analysis of Historical Data

• Managed a dynamic project involving the analysis of historical data and adaptation of decision-making processes in response to evolving Amazon product policies. Demonstrated flexibility and strategic thinking in navigating regulatory changes and ensuring compliance with updated guidelines.

Parity Row | Conversion of Non-English Defective Catalogue Products for English Marketplace

• Led efforts to identify and replace defective and suspicious keywords with English equivalents for improved accuracy in Amazon catalogue listings. Implemented machine learning algorithms to facilitate keyword detection and replacement.

Fuzzy Migration

• Analyzing keyword variations using machine learning algorithms to ensure rejection of defective ads from non-defective ones and implemented rules ensuring precision above 90%.

Wiki (SOP) scrapping

• Identifying high flagging defective keywords and forming combinations for ads title and OCR by implementing algorithms to detect defective ads and auto reject saving human moderation hours.