The challenge

Relevance AI helps companies build AI agents that can handle everything from customer support to sales outreach. These agents rely on the company’s advanced Retrieval-Augmented Generation (RAG) system to deliver fast, accurate responses.

But as demand for its platform surged, Relevance AI’s in-house vector database—the backbone of its RAG system—struggled to keep up. It couldn’t deliver the ultra-fast performance needed for customer-facing AI automation, creating bottlenecks that impacted the speed and reliability of AI agents. “Every millisecond counts, and slow vector searches were limiting our AI agents from delivering instant, accurate responses,” says Jacky Koh, Co-Founder and CEO of Relevance AI.

Maintaining and optimizing vector search was also draining internal resources. To sustain its rapid growth and meet customer expectations, Relevance AI needed a solution that could deliver fast vector retrieval at scale and keep AI agents running smoothly.

Fast, efficient vector search for unstoppable AI agents

Relevance AI turned to Redis to power high-speed vector search with sub-millisecond latency, allowing AI agents to retrieve relevant information and generate responses instantly. After testing multiple solutions, including Qdrant, OpenSearch, and their in-house implementation, Relevance AI found that Redis outperformed them all. Redis’ enterprise-grade vector search delivered speed and scalability, improving response times and semantic caching—capabilities that Relevance AI’s in-house solution couldn’t achieve on its own.

With Redis, Relevance AI optimized its RAG system for real-time performance. Faster vector search meant AI agents could now operate autonomously, retrieving, processing, and generating responses without delay.

By integrating Redis, Koh and his team drastically improved vector indexing, letting AI agents run on autopilot. With Redis’ caching, search times dropped by 99.5%—from two seconds to ten milliseconds—keeping AI responses instant and seamless at every search opportunity.

Scaling AI agents for streamlined sales & customer support

Relevance AI’s customers rely on the company’s AI agents to handle sales and customer support more efficiently. With Redis-powered vector search, these agents can quickly answer customer inquiries, freeing human teams for higher-value work.

In customer support, AI agents handle customer questions, resolve common issues, and escalate complex cases as needed, reducing response times and improving resolution rates. For sales teams, AI agents automate lead qualification, outreach, and CRM updates. With Redis’ low-latency vector search, AI agents can instantly analyze sales data, personalize prospect engagement, and close deals faster.

Redis-powered vector search hasn’t just improved the AI agents that Relevance AI’s customers use—it has also transformed the company’s own sales and customer support functions. Internally, Relevance AI has fully automated its SDR function using Redis-powered AI agents. No more manual prospecting, crafting cold emails, and logging interactions—AI agents now manage 100% of sales outreach and CRM updates in real time. This means sales teams can focus on closing deals rather than chasing leads.

By combining Redis’ fast vector search with advanced AI automation, Relevance AI has created a sales process that’s more efficient, scalable, and effective—both for its customers and within its own business.

Automating marketing workflows for faster execution

Beyond sales and customer support, Relevance AI has expanded AI automation to marketing, using Redis-powered agents to streamline campaign outreach, CRM data management, and procurement processes. By automating these workflows, AI agents help marketing teams execute initiatives faster and more precisely.

“We’re using AI agents to handle more of the tedious, repetitive tasks—whether it’s updating CRM data or completing long procurement forms,” says Koh. “The more you work with LLMs, the more you see the potential to streamline workflows and free teams to focus on higher-impact work.”

With Redis, Relevance AI has made its AI solutions faster and more reliable—both internally and for customers. Context-aware responses improve the UX, in a fraction of the time it’d take to achieve similar results manually.

Empowering today’s workforce with smarter AI service agents

With Redis, Relevance AI has automated key business functions for both its customers and internal teams. Redis-powered AI agents can quickly handle routine sales, customer support, and marketing tasks, allowing human employees to focus more on strategic work.

Redis’ speed and scalability allow the company’s AI agents to operate efficiently and reliably, even as the platform expands. From qualifying sales leads to resolving customer issues, Redis’ vector search capabilities allow agents to deliver fast, accurate answers.

As Relevance AI continues to grow, Koh expects Redis to play a key role in driving the next generation of intelligent automation. Redis’ real-time processing and scalability make it perfect for powering advanced AI workflows. “We’re continuing to explore innovative uses for Redis, including semantic memory routing and multi-agent systems,” says Koh. “We see Redis powering our future vision.”