Mesa Liquidity Data Generator

Learning objectives

• Explain how customer balances, account mix, and transaction behavior feed into a bank liquidity view.
• Generate synthetic data first, then compare the static liquidity snapshot with the Monte Carlo forecast.
• Interpret the most likely liquidity path, risk buckets, and AI-style coaching as decision support for treasury action.
• Discuss why simulation output is still only a model and should be checked against policy limits and stress assumptions.

Run modes

• Supported modes: Browser, Local
• Demo type: Multi-mode demo
• Primary launch surface: index.html

Before you start

Expect local setup time for Python dependencies before the first run, then faster repeat execution.

This helps set classroom expectations before students click into the live experience.

Core context

Look for the main decision, data input, and output the demo is designed to explain.

Observe how changing one assumption changes the result or recommendation.

Concepts covered

What students should note

Note the limitation students should mention before applying the result in a real decision.

List of steps

• Choose the run mode that fits the class: Browser, Local.
• Review the default assumptions before changing anything.
• Change one or two inputs, then use `Generate Data`.
• Read Monte Carlo Simulation first, then compare any supporting metrics, charts, or AI text.
• Capture one insight, one limitation, and one action recommendation.

Input variables explained

• `Enable AI-style insights` adds classroom-friendly commentary on the generated liquidity picture and forecast output.
• Use the generated customer, account, and transaction views as the base dataset before discussing the forecast result.

Decision buttons explained

• `Generate Data` creates the synthetic banking dataset that the later tabs and metrics depend on.
• `Run Monte Carlo Simulation` projects many future liquidity paths so students can see uncertainty rather than a single deterministic answer.
• `Export CSV` downloads the synthetic dataset for spreadsheet or Python follow-up work.
• `Reset` returns the demo to a clean teaching state.
• The `Overview`, `Customers`, `Accounts`, and `Transactions` tabs help students move from raw data to liquidity interpretation.

Outputs explained

• `Liquidity metrics` summarize the current synthetic balance-sheet position before stress is introduced.
• `Monte Carlo Simulation` shows the forecast distribution, progress, and most likely liquidity outcome under repeated simulated paths.
• `Liquidity coaching` translates the numbers into a treasury-style explanation of buffer strength and potential pressure points.

What to notice

• Compare the current liquidity snapshot with the simulated future distribution instead of treating them as the same thing
• Watch how the most likely value can still sit beside a wider risk range that matters for funding and escalation decisions
• Use the customer, account, and transaction tabs to explain why the forecast behaves the way it does

Discussion and reflection

• What business or technical decision would you make differently after using Mesa Liquidity Data Generator?
• If you changed one assumption and ran `Generate Data`, which output moved the most and why?
• What would you still want to validate with real data, policy, or expert review before acting on the result?