Morpheus

Supply Chain Resilience Agent - this AI agent would proactively identify and mitigate risks within a supply chain, helping businesses avoid disruptions and ensure the timely delivery of goods.

Comment

Inspiration

Global supply chains are increasingly vulnerable to disruptions - from geopolitical tensions to natural disasters. We needed an AI agent that could autonomously monitor, analyze, and predict supply chain risks in real-time, going beyond static dashboards to provide proactive intelligence that saves businesses millions in disruption costs.

What it does

Our Supply Chain AI Agent provides: • Real-time tracking of flights and vessels worldwide using live APIs • Autonomous risk analysis combining geopolitical events, weather, and logistics data • Predictive intelligence that proactively identifies potential disruptions • Multi-step reasoning to orchestrate complex supply chain scenarios • Interactive chat interface for natural language supply chain queries

How we built it

• AWS Bedrock Agent with Claude 3.7 Sonnet as the reasoning engine • 5 action groups handling tracking, risk analysis, search, and orchestration • 8 Lambda functions processing real-time data from 6+ APIs (AviationStack, AISStream, SerpAPI, NewsAPI) • React frontend with sleek dark UI deployed on S3 • API Gateway connecting frontend to Bedrock Agent • Real-time integrations for live flight/vessel tracking and news intelligence

Challenges we ran into

• API rate limits - solved with intelligent fallback mechanisms across multiple data providers • Real-time data synchronisation - implemented robust error handling and retry logic • Complex multi-tool orchestration - designed autonomous workflows that combine multiple data sources intelligently • Cost optimisation - balanced powerful AI capabilities with free-tier AWS services

Accomplishments that we're proud of

• Fully autonomous AI agent that makes real decisions without human intervention • Live vessel tracking using actual AIS maritime data (successfully tracked SUNRISEWAY crude oil tanker) • Enterprise-grade architecture with 100% AWS services and production-ready scalability • Sub-$1 monthly cost while delivering enterprise capabilities • Real-time intelligence combining 6+ data sources into actionable insights

What we learned

• AWS Bedrock Agent primitives enable sophisticated autonomous reasoning • Multi-API orchestration requires intelligent fallback strategies • Real-time supply chain data is incredibly complex but valuable when properly synthesised • Claude 3.7 Sonnet provides excellent reasoning capabilities at cost-effective pricing • Autonomous agents can genuinely replace human decision-making in specific domains

What's next for Supply Chain AI Agent

• Predictive modelling using historical disruption patterns • Integration with ERP systems (SAP, Oracle) for enterprise deployment • Mobile app for supply chain managers on-the-go • Advanced ML models for demand forecasting and route optimisation • Multi-tenant SaaS platform serving multiple enterprises • Blockchain integration for supply chain transparency and verification

Built With

Share this project:

Updates