Get a clear picture of how your team is performing across conversations. The new dashboard gives you full visibility into the metrics that matter most—response times, resolution rates, and how workload is distributed across your team.

Whether you're managing a growing support team or overseeing multiple agents, you now have the data you need to make smarter decisions, faster.

With the new in AgentFlow, you can connect external systems through a split-screen AI copilot that walks you through setup in plain language.

Upload your API documentation in JSON or Swagger format, and the copilot reads it, interprets the setup details, and maps everything automatically — no manual field-by-field work required.

Once connected, you can:

A practical example: a retail brand can connect its order system so the Agent looks up real-time order status when a customer asks "where's my package?" — pulling the answer directly from the source instead of handing off to a human agent.

The result: your Agent does more on its own, your team builds richer automations without engineering bottlenecks, and you unlock use cases that used to be off the table.

Until now, uploading source files to your AI Agent was a black box. You couldn’t see how documents were processed, what your Agent pulled from them, or where to fix issues when something felt off.

The new wiki-style Knowledge Base changes that. Uploaded sources are now transformed into you can review, edit, and assign — giving you full visibility into the knowledge powering every conversation.

Here's what you can do:

The result: your team can resolve knowledge conflicts with confidence, and keep your Agent's context accurate over time.

Team members can view, create, edit, and delete tickets directly from within a conversation — enabling faster issue tracking and resolution on the go without needing to switch to the web app.

When a customer shares new information during an AI conversation — like an updated phone number, email, or interest level — that data is lost unless someone manually updates the record.

Now, with the new actions in the Playbook step editor, your AI Agent automatically captures what the customer shares and syncs it directly to their SleekFlow Contact or Data record in real time.

When you select this action, a searchable dropdown lets you map the right property — so the agent knows exactly where to save what it hears. The update only fires when that specific Playbook step is triggered, so nothing gets overwritten unexpectedly.

Your team stays focused on the conversation, and your contact data stays accurate without anyone lifting a finger.

With the new Shopify actions in the Playbook step editor, your AI Agent can now guide shoppers through their buying journey by performing real commerce actions directly within the conversation:

Picture a customer asking, "Do you have this dress in medium?" — your AI Agent can find the product, confirm the variant, drop it into their cart, and share the checkout link, all without a human agent stepping in.

Your team gets to focus on the conversations that really need a human, while customers get faster, more complete support that moves them closer to checkout.

Sometimes when a customer reaches out, your AI agent may ask for a name, order number, or preference that your team has already saved in their profile. This can make the experience feel impersonal.

Now, with the new block in your Agent settings, you can choose exactly which properties your AI Agent is allowed to read — things like name, email, membership tier, or past purchase history.

Those details are automatically woven into the conversation behind the scenes, so your AI Agent can greet customers by name, skip questions you already have answers to, and tailor its responses to who it's actually talking to.

You stay in control of what gets shared: selected properties appear as chips in the configuration view so you can see at a glance what your Agent has access to — nothing hidden, nothing unexpected.

The result is a conversation that feels less like a support ticket and more like talking to someone who actually knows your customer.

When your AI Agent sends one long block of text, it can feel robotic and overwhelming for your customers.

Now, with in AI Agent Settings, you can toggle on a mode where your Agent automatically breaks longer responses into shorter, sequentially delivered chat bubbles with natural typing pauses in between — just like how a real person messages.

A few things worth knowing about how it works:

The result: conversations that feel more like talking to a person, and replies that are easier for your customers to read and act on.

Setting up an AI agent used to mean jumping between AgentFlow and Flow Builder. Now, the new AI Agent Setup walks you through the whole process in three steps: Create, Configure, and Deploy.

In the first Create step, the new template gallery gives you a head start. Instead of starting from scratch, you choose a preconfigured template with the Playbook instruction already mapped out, then customize it to make it your own.

The Configure step is where your agent takes shape. You can:

The Deploy step lets you assign your agent to one or more channels, set its active hours, and define what happens when it needs to hand off — whether that's a timeout or a customer asking for a human. Each exit condition can trigger follow-up actions like assigning to a team member.

You can now create, view, edit, and delete teams directly from the SleekFlow Settings. Manage team members, assign dedicated channels, and configure QR code settings—all in one place, without switching back to SleekFlow 1.0.