Signify, a global expert in connected LED lighting, faced a big challenge: answering complex technical questions from customers quickly and accurately. With thousands of different products and intricate details, their knowledge system was slowing down customer support. To fix this, Signify teamed up with Microsoft Research Asia to test a new AI-powered solution called PIKE-RAG. This innovative technology helped improve answer accuracy and made customer service more efficient.

Challenges of Using AI in Lighting Support

Signify understood that having precise and fast knowledge systems was crucial in today’s AI-driven world. They adopted large AI models and retrieval-augmented generation (RAG) techniques to better handle customer inquiries. But applying RAG in lighting support wasn’t simple. The product data included a mix of documents, unstructured tables, and detailed component parameters. This made customization slow and limited how well the system could grow and adapt over time.

Because of this, Signify needed a more advanced approach that could handle the complexity of their data. They wanted a system that could process different formats, understand technical content, and provide reliable answers to professional users. That’s when they started exploring solutions like PIKE-RAG, which had already shown success in fields like healthcare and law, where accuracy is critical.

How PIKE-RAG Solved Signify’s Problems

Partnering with Microsoft Research Asia, Signify ran a proof-of-concept (PoC) using PIKE-RAG on Microsoft Azure. This new system could efficiently search through textual data and understand multimodal content, such as charts and diagrams. It also quickly learned reasoning patterns specific to the lighting industry, making its responses more relevant and accurate.

One key feature of PIKE-RAG is its ability to adapt to a domain quickly. It learns from the data it processes, which means it gets better at understanding complex questions over time. For Signify, this meant that the AI could handle multimodal documents like circuit diagrams and nonstandard tables, providing organized and precise answers that customers could trust.

This upgrade resulted in better customer support, with faster and more reliable responses. The system’s improved understanding helped Signify deliver a higher level of service, giving them a competitive edge in the market and making customers happier with their technical support experience.

In summary, Signify’s partnership with Microsoft Research Asia and the integration of PIKE-RAG technology brought significant improvements. It addressed the unique challenges of lighting product data and transformed how the company supports its customers. As a result, Signify now offers more accurate answers, quicker responses, and a smoother support experience that keeps them ahead in the industry.

Inspired by

• https://www.microsoft.com/en-us/research/blog/when-industry-knowledge-meets-pike-rag-the-innovation-behind-signifys-customer-service-boost/

Sources

• signify.com
• azure.com
• github.com
• aka.ms