Course Outline
Session 1: Foundations and Prompt Engineering
Instructional Segment (45 min):
- Comparing Claude, ChatGPT, and Copilot for investment, finance, and analyst applications
- Applying the RACEF prompt framework (Role, Action, Context, Exemplars, Format) to investment analysis, forecasting, and model commentary
- Distinguishing between areas where model output can be trusted and areas requiring verification, along with basic data hygiene principles
Practical Exercise/Demonstration (20 min):
-
Each participant rewrites a typical prompt using the RACEF framework and compares the resulting outputs
Session 2: Financial Modelling, Long Documents, and Claude Code
Instructional Segment (45 min):
- Generating forward forecasts from sanitized management accounts, with Claude explaining the underlying assumptions
- Extracting waterfall structures from sanitized legal documents (the 'legal Bible' workflow)
- Introducing Claude Code for advanced Excel tasks, scripted analysis, and automating repetitive model activities
Practical Exercise/Demonstration (20 min):
-
Hands-on activity: converting a legal document into a waterfall template, followed by a Claude Code demonstration on a sample financial model
Session 3: Skills, Artifacts, and Claude Design
Instructional Segment (45 min):
- Skills: Developing reusable workflows (e.g., forecast builder, IM summariser, waterfall extractor)
- Artifacts and Projects: Managing shared, editable, and persistent outputs across the team
- Introducing Claude Design for creating charts, slide-ready visuals, and branded documents
Practical Exercise/Demonstration (20 min):
-
Each participant creates one reusable skill and generates one Claude Design artifact linked to a recurring task
Session 4: Context Windows, Claude Cowork, and Implementation Planning
Instructional Segment (45 min):
- Understanding 'context' and how its limitations manifest in long IMs, models, and multi-day analyses
- Strategies for chunking, summarisation, and project-level context management
- Introducing Claude Cowork for agentic, multi-step tasks that maintain momentum across complex workflows
Practical Exercise/Demonstration (20 min):
-
Demonstration of Claude Cowork executing an end-to-end task, followed by each participant drafting a personal implementation plan (for this week, this month, and this quarter)
Requirements
- Participants are expected to be engaged in investment, finance, corporate finance, private equity, M&A, FP&A, valuation, research, or analyst functions.
- A foundational understanding of financial models, management accounts, forecasts, investment memos, or transaction-related documents is assumed.
- Proficiency in Excel and experience with finance-specific document reviews are advantageous.
- Programming expertise is not a prerequisite, though familiarity with structured Excel models will enhance the experience during the Claude Code demonstration.
- Prior experience with Claude, ChatGPT, Copilot, or comparable AI tools is beneficial but not required.
Testimonials (1)
Greater knowledge on the topic
