Abhishek Mazumder

WORK HISTORY

EDUCATION

ABOUT ABHISHEK MAZUMDER

Abhishek is an analytics and AI professional with 15 years of industry experience. He is engaged in initiatives that are poised to change the face of the healthcare insurance business by being part of a team which is focussed on leveraging Gen AI to build Agents for propelling automation using cutting-edge Large Language Models (LLMs), resulting in a remarkable improvement in customer centre experience & claim processing lifecycle. To streamline manual efforts, reduce average customer handling time, expedite processing times, and boosted documentation processing leading efficiency in the Claims Processing domain to improve NPS scores and Advocate experience. Utilizing State-of-the-Art Prompt Engineering Techniques, RAG, NLP and other AI/ML techniques to facilitate cognitive search, pattern discovery, data extraction, and content generation essential for automation through plugable endpoints.Currently building agents around auto claim processing using LLM\'sI spearheaded the development of a Customer Call Summarization project, automating post-call documentation within the customer care center. Leveraging LLMs, this solution extracted key insights from customer interactions, expediting documentation processes while maintaining accuracy and consistencyIn collaboration with Business Leadership, I implemented a benchmarking framework for the performance metrics of the LLM like Chat Gpt 3.5/4 or Open source LLM like Llama 2/3 model against business defined KPIs and qualitative indexes. Used Stakeholder feedback to refine the performance, by few-shot prompting using gathered ground truths.Working on Information Retrieval techniques, to extract insights from heterogeneous Data Silos, enriching knowledge repositories using RAG and leveraging large context length of new born LLM\'sAdditionally Implementing Voice Based Question Answering Bot for Inbound Customer Care, We used best in business Automatic Speech Recognition service with Transformer-based Natural Language Understanding techniques. Integration of Entity Recognition, Part-of-Speech tagging, Topic Modelling, and functionalities to ensure unparalleled customer engagement and satisfaction.Also worked on implementing responsible use of AI framework developed within the organisation in our auto summarisation solution to make sure that there is no bias in the Gen AI solutions that are implemented in production.US B1/B2 Visa holder till `28.