"Generative AI for Finance: Building Agents for Research and Trading" Generative AI is rapidly reshaping how financial institutions research markets, build strategies, and execute trades. This book is written for quantitative researchers, data scientists, technologists, and forward-looking portfolio managers who want to move beyond toy demos and build robust agentic systems for real-world finance. Blending practical engineering with rigorous quantitative thinking, it shows how to turn large language models into reliable collaborators for idea generation, analysis, and decision support. You will learn how to construct end-to-end pipelines that feed high-quality market and textual data into transformers, retrieval systems, and agent frameworks. Core topics include prompt design for financial tasks, retrieval-augmented generation over filings and news, and orchestration patterns that let agents reason, plan, and call tools such as risk engines, databases, and execution venues. The book also develops skills in evaluation, backtesting, and MLOps so that research and trading agents are measurable, debuggable, and governable rather than opaque black boxes. The text assumes comfort with Python and basic statistics but reviews the essentials of quantitative finance, machine learning, and reinforcement learning as needed. Examples emphasize production-grade design-covering performance engineering, compliance, security, and model risk-making this a practical guide for building agentic AI systems that can survive contact with real markets and real regulators.