Week of September 18th
New updates include: Policy Builder
Features
Introducing Policy Builder - Welcome to true conversation orchestration
LivePerson is pleased to announce the launch of Policy Builder, our real-time actions-based platform within Conversation Cloud.
What’s fueling our excitement for this? Policy Builder autonomously supervises your customer conversations, ensuring your strategic rules are enforced. Just how? It monitors all automated and human conversations in Conversational Cloud and—when important situations arise—it intervenes in a timely manner to achieve specific outcomes.
Policy Builder listens for certain situations or “signals,” such as a financial distress message, an attempt to hijack the bot, or simply that a conversation has been closed. Several signals are available out-of-the-box, and you can register your own custom signals too (Bring Your Own Signal). Policy Builder also supports a broad spectrum of resulting actions, ranging from routing to workflow execution.
Within Policy Builder, define the business policies that are important to your brand, and apply them to bots and/or skills. A policy is the definition of one signal plus one or more rules (sets of actions).
Creating a new policy
An example policy
Policy Builder revolutionizes the way that you engage with your consumers. Use it to supervise every conversational interaction with responsiveness, adaptability, precision, and consistency.
Welcome to the era of conversation orchestration.
Key capabilities
- A policy-based system for orchestrating conversations between consumers and your human agents, LivePerson bots, and third-party bots.
- Out-of-the-box support for content moderation and AI Safety use cases.
- Out-of-the-box support for popular signals, such as conversation end, consumer sends message, and agent or bot sends response.
- No-code to low-code rules and actions.
- Support for custom signals via Bring Your Own Signal (BYOS).
- Support for custom actions via Bring Your Own Action (BYOA) - Integrate with any REST-based web service or API.
Key benefits
- Uber control: Policy Builder can override an agent or a bot in any conversation at any time.
- Operational efficiency and reduced complexity: Policy Builder significantly reduces the development and operational efforts traditionally required to supervise consumer interactions. It simplifies the process, reduces the risk of errors, and eliminates the complexities associated with manual interventions, thus freeing up resources and reducing operational costs.
- Proactive security measures: Policy Builder enhances safety by proactively monitoring conversations for any unusual activity that might indicate security threats, such as attempts to hijack a bot or unauthorized access attempts. By detecting these signals early, it allows you to take immediate action to mitigate risks, safeguarding both your company’s data and that of your consumers.
- Quick and easy deployment of advanced automations: Streamline your Conversational AI solution by defining policies that apply to multiple LivePerson bots or agents. Got 10, 20, 50 bots or more? Apply a single policy to them all. This greatly simplifies the implementation of each bot, and it significantly eases your bot maintenance burden.
- Low-code setup: Engage your business users with creating policies.
- Customize and extend: Leverage or extend the platform to support custom use cases.
Limitations
Currently:
- You can’t create policies that apply to voice bots or post-conversation survey bots.
- Support for languages other than English is experimental only. That is, the conversations that you want to monitor and manage, the policies that you want to create, the intent domain that you rely on—everything—must be in English. As always, proceed with care: Test thoroughly before rolling out to Production.
Interest piqued?
If you have a use case for Policy Builder that you want to discuss, please schedule a call with our product team using this link.
Coming soon!
Stay tuned for our launch and more detailed documentation in the Developer Center.