The American Red Cross (ARC), a leading humanitarian organization, partnered with Dimension Labs to enhance the performance and effectiveness of their chatbots—an essential tool for scaling emergency assistance and disaster relief efforts. By utilizing Dimension Labs’ data platform, ARC optimized its chatbot interactions, enabling faster response times when vulnerable populations needed help the most.

When disaster strikes, we have to be there for others during their time of need. Technically speaking, we have to quickly interpret conversations at scale and adjust to new requests and conversation topics. Dimension Labs gives us the insights to provide the support needed to those who need it most.

Additionally, I now monitor the performance of our engagement channels that outside contractors manage. The ability to proactively monitor saves me time and resources. — Hailey Burgess, Product Manager

Key Metrics: 

• Improved disaster response goal attainment (e.g., ensuring individuals seeking shelter could successfully find it).
• Lowered not-handled/fallback intent, with a particular focus on improving Spanish-language interactions.

The Challenge:

The ARC faced critical challenges in delivering timely and effective disaster relief, including:

• Difficulty understanding user behavior during high-volume periods, such as natural disasters, without real-time insight into chatbot interactions.
• Manual review of chat transcripts, which was time-consuming and insufficient for scaling disaster response training.
• A lack of scalable processes to generate training data for improving natural language understanding (NLU) engines.

These issues created bottlenecks in the ARC’s ability to identify user priorities, fix problems quickly, and ensure critical resources were delivered to those in need.

The Solution:

Dimension Labs provided the ARC with an advanced data platform that streamlined and automated how they analyzed chatbot interactions during emergencies.

Key Capabilities Utilized:

• Topic Modeling and Behavioral Insights: Dimension Labs’ topic modeling feature quantified user behavior during high-volume periods, offering ARC a clear understanding of user needs in real time.
• Scalable Training Data Creation: Automated processes identified and extracted novel training data for ARC’s NLU engine, significantly reducing the time required for bot retraining.
• Real-Time Interaction Analysis: The platform delivered real-time visibility into chatbot interactions, helping disaster response teams identify and address user priorities more efficiently.

Dimension Labs’ AI-driven enrichment allowed the ARC to improve chatbot performance, even during peak demand, ensuring their bots could act as effective frontline resources during disasters.

The Results:

The partnership with Dimension Labs enabled ARC to:

• Improve disaster response goal attainment, ensuring individuals in need could find shelters and essential support during emergencies like Hurricane Ian.
• Lower fallback intent, especially in Spanish-language interactions, providing critical support to diverse populations.
• Shift to user journey-focused reporting, offering stakeholders a more comprehensive view of interactions and better tracking of KPIs.

Dimension Labs empowered the ARC to deliver faster, more reliable support when it mattered most, solidifying their chatbots as a cornerstone of their disaster relief strategy.