Zishaan is a Data Analyst experienced in designing and deploying AI/ML solutions. He has expertise in implementing RAG architectures using Langchain and optimizing chatbots with LLMs like Claude-3. His skills include document classification using fine-tuned models (LayoutLMv3) and advanced data extraction from various document types using AWS services.
Experience
Data Analyst
Ganit Business Solution
Implemented RAG architecture to on LLM model application that can have knowledge-base as custom data provided by the user. Optimized the chatbot built on Langchain and Claude-3 LLM for fast and more accurate output by retrieving responses from locally stored embeddings. The input can be any document (PDF,JPEG,PNG) which is passed through an OCR (AWS-Textract in this case) to generate the text and store them into VectorDB after applying Embeddings on int which user can access through its query. Fine-tuned the LayoutLMv3 model on custom data with 13 ID Labels including Adhar-Card, PAN-Card, Voter-ID. Used prompt advancement to get output in Key-Value pairs, with the use of model to detect the Names, Adhar No., DOB and address from the Adhar Card that is feed as image input.
Intern Data Analyst
Ganit Business Solution
Proficiently utilized Python(Pandas) and SQL to manipulate and analyze data, contributing to data-driven strategies. Assisted in data collection and preprocessing, ensuring data accuracy and completeness for analysis.
Education
MCT Rajiv Gandhi Institute of Technology
Bachelor of Engineering (B.E.)
Computer Science
Dr. DY Patil School
Higher School Secondary Certificate (HSC)
Science (Physics, Chemistry, Math)