Bret Taylor, former co-CEO of Salesforce and current OpenAI board chairman, has launched the new AI agent startup, Sierra, to “transform the customer experience”.
Describing itself as the “conversational AI platform for businesses”, Sierra intends to empower enterprises to build their own distinct AI-powered agents and elevate their digital operations. These agents can process customer conversations at a large scale, allowing time and effort for human agents to focus on higher-priority tasks.
Sierra has already raised $110 million in funding from several investors, such as Sequoia and Benchmark, and is cofounded with Google’s former Head of VR, Clay Bavor.
In its announcement blog, Taylor and Bavor wrote:
At Sierra, we’re building the conversational AI platform for businesses, enabling every company – including yours – to build their own agent(…) Our enterprise-grade platform is powerful, easy to deploy, and capable of creating AI agents that are sophisticated, authentic, and trustworthy.”
Sierra has already attracted multiple large businesses as customers, with its website featuring testimonies from Weight Watchers, Sonos and OluKai. These companies credit Sierra’s technology with allowing them to “scale and reach more customers, all while using our voice, delivering white glove service”, according to Kerry Konrady, Chief Marketing Officer at Hawaiin-inspired clothing brand Olukai.
Taylor left Salesforce in January 2023 and then received a call from Bavor, who he worked with at Google. According to a Fortune interview with the Sierra co-founders, their shared interest in AI’s transformative potential inspired their decision to start an AI business together.
“I looked at these amazing technologies and it’s so easy for the technology companies around us to deploy because we’ve got buildings full of engineers who can follow the latest research,” Taylor told Fortune. “The greatest opportunity we have is to enable every company, no matter how sophisticated or technical, to deploy (AI) successfully.”
- OpenAI Has Launched a GPT Store. Here’s What It Means for Contact Centers
- Salesforce CEO Marc Benioff Cautions OpenAI on Its “Ripped Off” Training Data & “Commoditized” Models
Sierra’s Platform Feature Set
Sierra’s agents can integrate with the business’s legacy infrastructure to access company data and records, which can then inform actions — if formally approved. This process enables agents to perform an even greater number of tasks than conventional conversational AI agents, such as understanding and responding to the nuances of an issue in the order management system to effectively oversee customer subscriptions.
Sierra’s ambition is to revolutionise the conversational AI space by parsing the complexity of human interactions, which even the most advanced chatbots sometimes struggle to achieve. These always-available agents can grasp technical terminology, typos, and the overall context of discussions.
“Agents can reason, problem solve and make decisions,” Taylor and Bavor wrote. “With Sierra, you set goals to guide your agent toward the right solutions and guardrails to ensure your agent stays on point and aligned with your policies. No workflow or process is too complex.”
Intriguingly, Sierra also promises that its agents will display genuine empathy for customer frustrations or concerns. They respond in the language preferred by users, showing care and empathy while adjusting to their unique requirements and emotional states.
Sierra stresses that while Sierra’s agents are informed by business policies, records and knowledge base, “those interactions are deterministic, ensuring your agent always adheres to your security policies and access controls”, its blog wrote. “The Sierra platform also enforces strict data governance, protecting your customers’ personal information and ensuring that your company’s data stays your own.”
The platform includes “powerful” auditing and QA tools so admins and managers can understand the reasoning behind every agent decision.
When a problem has to be escalated to the customer service team because the AI agent can’t address it, the agent provides the human agent with a “detailed summary” so they’re equipped to deal with the issue. Customer service teams can also view agent interactions in real time to immediately pick up flagged customer cases.