FreightAssist Agent

Inspiration Handling logistics-related queries—especially delayed shipments—is often a stressful and manual task for service agents. With data scattered across systems and limited visibility into contract terms, the customer experience suffers. We wanted to simplify this journey by empowering agents with a smarter, AI-driven solution that gives them instant insights and decision support. What it does FreightAssist is an AI-powered virtual agent designed to support logistics and shipment queries. It provides real-time shipment status, reasons for delays, and visibility into customer policy and contract documents—all in one place. It reduces system toggling and escalates to Salesforce only when needed. The solution is also integrated with WhatsApp and Slack for easy communication and collaboration. How we built it We leveraged a range of Salesforce technologies and external integrations to create FreightAssist:

• Agentforce to build the AI-powered support agent
• Data Cloud for unified, real-time access to customer and shipment data
• Prompt Builder to generate intelligent and context-aware responses
• WhatsApp Integration using Apex and external connectors for direct customer interactions
• Slack Integration for internal team collaboration and rapid agent support
• Apex for logic, workflow automation, and Salesforce case creation
• Custom APIs for real-time shipment tracking and delay insights
• A user-friendly agent UI for quick navigation and faster resolution Challenges we ran into
• Consolidating fragmented data sources into a single, real-time view
• Handling edge cases related to policy exceptions and contract conditions
• Ensuring a consistent user experience across WhatsApp, Slack, and the agent UI
• Last-minute bugs and final presentation prep (big thanks to Aditya and Mudarasar!) Accomplishments that we're proud of
• Built a functional AI prototype in under 48 hours
• Achieved seamless integration with Salesforce, WhatsApp, and Slack
• Created a solution based on real-world use cases in logistics
• Collaborated cross-functionally, worked overnight, and delivered under pressure What we learned
• Good UX is essential for agent productivity
• AI becomes truly effective when backed by unified, real-time data
• Teamwork and agility make innovation achievable—even on tight timelines
• Platform integrations (like Slack/WhatsApp) can enhance usability immensely What’s next for FreightAssist
• Extend to other logistics use cases like returns and claims
• Expand AI training with real-world scenarios
• Build a self-service customer version
• Showcase the solution across teams and clients as part of Nagarro's innovation portfolio

Built With

• agentforce
• apex
• datacloud
• promptbuilder
• slack
• whatsapp