Ada release notes

Earlier this month, we released a version of Processes that used natural language instead of blocks to build response flows for your customers. After assessing the performance of those natural language Processes, we've decided that the best way to automate your customers' resolutions is by making your AI Agent's performance as reliable as possible - so we're bringing back the block-based building experience.

Processes is now better and more flexible than ever, with more blocks available for you to use, and the ability to reference variables. For more information, see Create workflows with Processes.

We've introduced some new settings to customize your AI Agent's persona, so you don't have to control its behavior in your Guidance. This includes settings that control:

• How long you want messages to be
• How many steps you want your AI Agent to put in a message
• Whether your AI Agent mentions your company's competitors

To learn about all of your AI Agent's persona settings, see Customize your AI Agent's personality in automatically generated content.

We're excited to release the newest version of Processes, which allow you to train your AI Agent to solve problems that require multiple steps and logic.

Now, instead of using blocks, you can write out multi-step instructions for your AI Agent to execute, complete with conditional logic and referencing Actions you've already configured. This way, you can use the same language you would use training a human agent to dictate what your AI Agent does.

As part of continually improving the accuracy of our automated resolution classification model, we have made two exciting updates for generative clients:

• Your AI Agent now evaluates all of its conversations for automated resolutions, instead of just a sample.

• We've improved the way your AI Agent evaluates whether it provided relevant information grounded in your knowledge content, meaning its classifications are more accurate.

Going forward, all of the the data on automatic resolutions in your reports, and your conversation topics, will include all conversations.

For more information, see Understand and improve your AI Agent's automated resolution rate.

We've made it easier to protect your customers' sensitive information. Now, you can use regular expressions to create matching patterns your AI Agent or bot can use to find sensitive information like credit cards and redact it. Redacted information stays within the conversation, is never visible in transcripts or audible in recordings, and never gets sent to an external LLM, APIs, or handoffs to human agents.

For more information, check out our Help Docs:

• If you have a generative AI Agent, see Redact sensitive information.
• If you have a scripted bot, see Create a custom redaction.

We've added customization options for emails, so they can better match your company's branding:

• Modify the sender's name
• Add a custom header image (e.g., a logo or a banner)
• Add text and links in the footer

For more information, see Customize emails.

When you go into the Conversations view, you can now see new messages appear just a few seconds after they happen, without having to refresh the page. If you’re in a conversation that’s currently happening, you can click the New messages notification to scroll down to the newest messages.

For more information, check out our Help Docs:

• If you have a generative AI Agent, see Improve your AI Agent's responses by viewing past conversations.
• If you have a scripted bot, see View past chatter conversations.

We've made it easier to select a speaking voice for your Voice generative AI Agent or scripted bot!

• We've given friendlier names to the speaking voices available in Ada (e.g., Myra, Markus, Paola)
• We've also added a descriptor that appears beside the voice name that describes the character of the speaking voice (e.g., Mature, Confident, Friendly, Relaxed)

For more information, check out our Help Docs:

• If you have a generative AI Agent, see Choose a speaking voice for your AI Agent.
• If you have a scripted bot, see Choose a speaking voice for your bot.

Until now, if you've used multiple Ada APIs, you've had to create and manage different keys for each one. Now, you have the option to create one or more global keys that work across all of Ada's current and future platform APIs.

All of your existing API-specific keys will continue to work, and will still be specific to the APIs they were originally generated for. They'll be marked as legacy keys in your dashboard, so you can tell them apart from your global keys.

For more information, see Authorize API integrations in our Developer Docs.

We’re excited to announce that your AI Agent can now respond to email inquiries!

Your customers can start a conversation with your AI Agent by simply writing an email to your chosen support address (new or existing, e.g., [email protected]). They can then converse back and forth with your AI Agent to get their issues resolved.

Your AI Agent will leverage knowledge, Actions, and Guidance configured in Ada to automatically resolve your customers' inquiries.

If a customer issue can’t be resolved, the AI Agent will forward the conversation to your team, who can then assist the customer using the agent platform you already have in place.

Like all conversations your AI Agent has, you can audit its email conversations, and then provide Guidance to coach and improve the way it interacts with your customers, and its ability to resolve customer issues.

To learn more about email functionality and how to enable Email on your AI Agent, see Get started with email.