Skip to main content
Back to challenges
Challenge

Multi-Currency Liquidity Forecasting Agent using LlamaIndex and Mastra AI

Global banks like Ant International are deploying AI forecasting models to predict cross-border FX liquidity requirements across multiple treasury jurisdictions. Effective liquidity optimization prevents capital locking while mitigating foreign exchange risks. In this challenge, you will implement an intelligent retrieval-augmented forecasting agent using LlamaIndex and Mastra AI. Your agent will index multi-currency cash flow statements and historical rate movements, retrieving context to feed dynamic time-series projection algorithms.

CybersecurityHosted by Vera
Sign in free to start
Challenge brief

What you are building

The core problem, expected build, and operating context for this challenge.

Build an intelligent liquidity forecasting system using LlamaIndex and Mastra AI for multi-currency treasury forecasting.

Delivery guide

How work is evaluated

Evaluation

Evaluates liquidity prediction accuracy and actionable rebalancing recommendations.

Datasets

Challenge datasets

Public datasets, plus the private datasets you have access to.

Evaluation rubric

How submissions are scored

These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.

Dimensions
2 scoring checks
Binary
2 pass or fail dimensions
Ordinal
0 scaled dimensions
Dimension 1

shortfall_detection

Accurately flags whenever a currency reserve drops below zero

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Dimension 2

forecast_mape

Mean Absolute Percentage Error on predicted daily closing balance • target: 3.5 • range: 0-100

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Learning goals

What you should walk away with

  • Construct a LlamaIndex VectorStoreIndex over financial time-series summaries

  • Build workflow graphs in Mastra AI to trigger sub-agents for multi-currency settlement analysis

  • Calculate 7-day liquidity shortfalls with high statistical confidence intervals

  • Expose automated rebalancing recommendations via standardized REST/gRPC endpoints

Resources and assets

Reference links and supporting material

Dataset notes

Multi-currency bank account cash flow traces including cross-border wire events, spot trade executions, and regional buffer constraints.

Challenge setup

Evaluates liquidity prediction accuracy and actionable rebalancing recommendations.

Challenge input

JSON treasury ledgers across USD, EUR, and SGD

LlamaIndex

Structured indexer and retrieval layer optimized for financial documentation.

Mastra AI

Provides robust state-machine agent execution in TypeScript.

Evaluated output

JSON containing projected_shortfalls and recommended_transfers

Evaluation checks
  • Accurately flags whenever a currency reserve drops below zero
  • Mean Absolute Percentage Error on predicted daily closing balance • target: 3.5 • range: 0-100
Configured evaluation evidence
  • Benchmark: FinNLP-TreasuryForecasting
  • Forecast Mape target: 3.5
  • 1 public reference case
Configured runtime
  • Python execution harness
  • Python sandbox (unavailable on Versalist)

Summary shows the first task, up to three tools, and up to three evidence items. Open the technical recipe to view all configured items.

View technical recipe

Configured tools

Action Space
  • Llama Index · Required
  • LlamaIndex · Optional
  • Llama 3.2 · Optional

Evaluation contract

  • shortfall_detection · Weight 1
  • forecast_mape · Weight 1

Metrics

  • forecast_mape · Target 3.5

Benchmarks

  • FinNLP-TreasuryForecasting
  • general

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON treasury ledgers across USD, EUR, and SGD

    Output: JSON containing projected_shortfalls and recommended_transfers

    Execution harness: Python

Evaluation tools

  • LlamaIndex · RAG Framework
  • Mastra AI · Agent Workflow Framework

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Multi-Currency Liquidity Forecasting Agent using LlamaIndex and Mastra AI