Bunty Shah

About Bunty Shah

AI & Machine Learning Leader | VP at MSCI | Driving Business Transformation with LLMs | Winner of Multiple Hackathon Challenges. Passionate about pushing AI and ML boundaries, I lead transformative initiatives at MSCI. I spearhead the adoption of LLMs, LLMOps, Advance prompt engineering techniques, revolutionizing equity and ESG domains. My contributions include:1⃣ Large Language Models (LLMs): Implementing GPT3.5, GPT4, PaLM (Bison) at scale. Developing LLMOps framework for evaluation and deployment. Optimizing information extraction from diverse sources.2⃣ Advanced NLP Frameworks: Upgrading NLP infrastructure with scalable annotation tools and distributed computing. Employing MLOps for improved model performance.3⃣ Data Extraction and Validation: Pioneering entity extraction from key documents. Establishing data store for fundamental data extraction from reports. Addressing validation challenges.I excel in developing innovative AI products. Notable projects include: Text AI: Comprehensive ML NLP tool for text analysis. Supports BERT with BiLSTM, text classification, sentiment analysis, topic modeling, and entity recognition. Vision AI: Versatile ML tool for image processing. Offers image ingestion, pre-processing, augmentation, classification, segmentation, and object detection. Speech AI: All-in-one ML tool for speech recognition. Enables speaker diarization, acoustic modeling, language modeling, punctuation, and domain-specific audio training. Let\'s connect to drive AI innovation and leverage LLMs and advanced techniques for unparalleled business value!

Experience

1. Vice President Machine Learning Engineer Llm Llmops Data Scientist Ai Architect

   MSCI Inc.

   Apr 2021 — Present · Mumbai, IN

   Vice President at MSCI | Transforming Businesses with AI & Machine LearningAreas of Expertise: Leveraging AI and ML capabilities to drive business transformation in equity and ESG domain. Leading efforts in the adoption of large language models (LLMs) to revolutionize multiple use cases like data extraction, chatbot development (langchain), text classification etc. at scale using advance prompt engineering and fine-tuning techniques. Evaluating open source LLMs for specific use cases with fine-tuning capabilities.Key Contributions:1⃣ Large Language Models (LLMs)- Spearheading multiple use cases with GPT3.5, GPT4, and PaLM (Bison)- Developing LLMOps framework for efficient evaluation and faster production deployment- Extracting thousands of data points from diverse sources and optimizing information extraction.2⃣ Advanced NLP Frameworks and Standardization- Upgrading NLP infrastructure to incorporate scalable annotation tools- Harnessing distributed computing for large-scale training and inference- Employing MLOps techniques to continually improve model performance over time.3⃣ Data Extraction and Validation- Pioneering the extraction of key entities from documents, such as private placements, IPO data, and dividend information- Establishing a data store for systematic extraction of fundamental data points from annual and interim reports (e.g, balance sheets, cash flows)- Enabling validation through vendor sources and addressing Non-Comparable (NC) and Dual Vendor Validation (DVV) cases. Excited to drive innovation in the AI & Machine Learning space, incorporating the latest trends and best practices to create unparalleled business solutions.Let\'s collaborate to push the boundaries of AI and ML with state-of-the-art LLMs and advanced Prompt engineering techniques. Together, we\'ll drive innovation & deliver unmatched business value!

Education

• Mumbai University Mumbai

  Bachelor of Engineering - BE, Electrical, Electronic and Communications Engineering Technology/Technician

  2006 — 2010