Customer conversations that change a system
Sierra builds AI agents for businesses to interact with their customers. Its platform connects conversation to operations such as managing a subscription, tracking a delivery or processing a product exchange. The agent needs access to the relevant business system as well as a language model that can interpret the request. [1]
The company introduced its platform publicly in February 2024. Founders Bret Taylor and Clay Bavor described agents already operating for consumer brands, while opening the platform more widely. This records a product with deployed workflows, rather than a company whose only milestone was a funding announcement. [1]
Policy and inspection around an agent
Sierra's development process uses a specification for the customer journey, its policies and required APIs. Its Experience Manager lets a business review interactions and refine those instructions. These controls connect the behavior requested by a service team with the systems an agent needs to use. [2]
Sierra's implementation account describes agents as products that need a development cycle: defining a customer journey, connecting systems, supplying policies, testing and improving from production observations. These are methods for operating the application around a model; they do not establish a universal success rate across deployments. [2]
Sources
- Meet Sierra, the conversational AI platform for businesses
Sierra · Feb 13, 2024
- Shipping and scaling AI agents
Sierra · Jul 25, 2024