GenAI in Finance: Architectures, Forecasting, Agents & Multimodal Retrieval
October 4–5 & October 18–19, 2025 | Virtual | All times in ET
Overview
Financial institutions today are at the cutting edge of adopting Generative AI (GenAI) to transform forecasting, compliance, research, and client interaction. This 4-day intensive program is designed for AI developers, ML engineers, financial data scientists, and technology leaders who want to master the architectures, tools, and deployment strategies powering the next generation of GenAI applications in finance.
Through live expert-led sessions, deep technical labs, and interactive Q&As, you’ll learn to design, optimize, and deploy large language models (LLMs), Retrieval-Augmented Generation (RAG) systems, and agent-based architectures specifically for finance-critical workflows.
Why Attend
- Cutting-edge technical depth: Go beyond theory with practical insights on transformers, multimodal RAG, and agent pipelines.
- Finance-focused applications: Learn how to apply GenAI to forecasting, compliance, portfolio analysis, and risk management.
- Hands-on learning: Office hours, interactive Q&A, and real-world case studies ensure you can put concepts into practice immediately.
- Industry-ready outcomes: Build the skills to design, evaluate, and monitor GenAI systems that meet robustness, transparency, and regulatory standards in financial environments.
Curriculum Highlights
Day 1 — Transformer Fundamentals, Optimization Techniques & Prompt Engineering (Oct 4)
- Transformer architectures: self-attention, long-context models, tokenization & embeddings for text, time series, images, and audio.
- Efficient fine-tuning: LoRA, QLoRA, quantization, sharding, inference scaling.
- Prompt engineering for finance: Chain-of-Thought, Tree-of-Thought, structured reasoning over financial documents and numbers.
Day 2 — Time Series Transformers, Audio Pipelines & RAG Fundamentals (Oct 5)
- Time Series Transformers: forecasting, anomaly detection, synthetic data generation for stress testing.
- Audio pipelines: transcription, classification, and sentiment analysis for financial audio (earnings calls, investor Q&A).
- RAG for finance: architecture, vector databases, chunking, hallucination mitigation, debugging retrieval performance.
Day 3 — Multimodal RAG Systems, Advanced Retrieval & Evaluation (Oct 18)
- Multimodal RAG: integrating text, tables, filings, and images.
- Structured extraction with LLMs: grounding outputs for compliance and analysis.
- Advanced RAG architectures: Corrective RAG, Self-RAG, Fusion RAG.
- Evaluation best practices: dataset generation, troubleshooting, mitigation of RAG pain points.
Day 4 — Agent Architectures, Deployment & Monitoring (Oct 19)
- Financial agents: LangGraph, ReAct, LATS, reflection, planning, tool-calling for multi-step reasoning.
- Tool-using agents: portfolio analysis, compliance flagging, chart-reading, code debugging.
- Deployment & monitoring: secure Streamlit pipelines, LangSmith & LangFuse observability, sandboxing, transparency, and guardrails in regulated settings.
Learning Outcomes
By the end of this course, participants will be able to:
- Architect, fine-tune, and deploy transformer models across modalities (text, time series, audio, image).
- Apply PEFT, quantization, and sharding to build efficient and deployable models.
- Prompt LLMs for structured reasoning in financial tasks using CoT, ToT, and advanced prompting methods.
- Forecast and generate synthetic time series data for financial modeling.
- Transcribe, classify, and analyze financial audio data for actionable insights.
- Design and optimize multimodal RAG systems for finance-specific use cases.
- Build, debug, and deploy agent-based systems for compliance, risk analysis, and portfolio optimization.
- Monitor and evaluate GenAI deployments for robustness, transparency, and regulatory compliance.
Who Should Attend
- AI/ML Engineers & Data Scientists building financial models.
- Quant Developers & Risk Analysts integrating AI into trading, forecasting, and stress testing.
- Engineering Leaders & CTOs seeking scalable, compliant GenAI deployment strategies.
- Financial Researchers & Strategists aiming to leverage AI for document analysis, audio reasoning, and portfolio intelligence.
Logistics
- Format: 4-day live virtual course
- Dates: October 4–5 & October 18–19, 2025
- Timings: 8:30–15:30 ET (with breaks)
- Interactive Office Hours included daily
