Suman Kumari

Senior Consultant @EY

IN

AI/ML Scientist | PySpark developers | Ernst & Young (EY)My responsibilities include:PDF Data Extraction: I extract unstructured data from complex financial documents using Python libraries such as Camelot, PyMuPDF (Fitz), and pdfplumber, as well as cloud-based tools like Azure Form Recognizer and Azure AI Document Intelligence. These solutions enable scalable ingestion of financial data into downstream analytics and RAG workflows.Data Extraction & Vectorization: I manage end-to-end RAG pipelines—processing financial content, converting it into vector embeddings, and enabling semantic search via vector databases to support intelligent querying and document retrieval.Process Automation with Generative AI: I use generative AI to automate financial operations such as reconciliation, automated commentary/narration, and report summarization, helping business teams reduce manual efforts, improve accuracy, and streamline reporting.Machine Learning & Deep Learning for Finance: I have built finance-focused models using CNN and RNN architectures for deep learning tasks. I have also implemented XGBoost, SVM, and Random Forest models for structured data problems like risk scoring, credit behavior prediction, and financial forecasting.Natural Language Processing (NLP): I have applied NLP techniques for text generation, document classification, and building intelligent assistants using large language models (LLMs), integrating them into RAG systems to enhance financial document interpretation and analysis.Big Data Processing with PySpark: I utilize PySpark to process and analyze large volumes of financial data efficiently, including feature engineering and transformation pipelines for model development.Cloud Computing & DevOps: I work across AWS and Azure platforms. On AWS, I’ve used EC2, S3, Lambda, and DynamoDB for scalable data workflows. On Azure, I build CI/CD pipelines using Azure DevOps, and design data and ML workflows using Azure Databricks.