[go: up one dir, main page]
Skip to main content

AI in Slack apps overview

There are multiple ways to engage with agents in Slack. This guide demystifies AI features in apps and agents from a developer's perspective.

What is an agent?​

AI Agents are autonomous, proactive applications designed to execute specialized tasks to help employees and customers. Using large language models (LLMs) to analyze and understand the full context of customer interactions or an automated trigger, they reason through decisions on the next steps autonomously. Agents can be instructed to generate responses that are consistent with your company’s brand voice and guidelines using trusted business data sourced from your CRM, Slack, and external applications. They are capable of operating 24/7, handling tasks proactively within set guardrails. When faced with complex issues beyond their scope, agents can escalate the matter to a human counterpart.

Use cases​

Agents can solve a variety of use cases in any industry. Knowing that the Salesforce Data Cloud, AI models, Slack, and third-party integrated apps are your data sources, the world is your agent's oyster. Take for example these use cases for agents across service, sales, marketing, and commerce:

  • Service Agent replaces traditional chatbots with AI that can handle a wide range of service issues without preprogrammed scenarios.
  • Sales Development Representative (SDR) engages with prospects 24/7, answering questions, managing objections, and scheduling meetings based on CRM and external data.
  • Sales Coach provides personalized role-play sessions for your sales team, using Salesforce data and generative AI to help sellers practice pitches and objections tailored to specific deals.
  • Merchandiser assists your ecommerce merchandisers with site setup, goal setting, personalized promotions, product descriptions, and data-driven insights, simplifying daily tasks.
  • Buyer Agent enhances the B2B buying experience, helping your buyers find products, make purchases, and track orders via chat or within sales portals. Personal Shopper acts as a digital concierge on your ecommerce sites or messaging apps, offering personalized product recommendations and assisting with search queries.
  • Campaign Optimizer automates the full campaign lifecycle, using AI to analyze, generate, personalize, and optimize marketing campaigns based on business goals.

How to engage with agents​

Depending on your use case and needs, you may choose to use a third-party agent or develop one yourself—also known as a first-party agent.

Third-party agents (available in the Slack Marketplace) are made by Slack partners and leverage AI in Slack with out-of-the-box functionality.

Developing a first-party agent gives you more flexibility with how you implement AI with a custom app in Slack. You will choose and implement your own LLM or internal-only database with this option.

In terms of Salesforce and Slack, there are two ways to create a first-party agent:

  • Build an agent in Salesforce and deploy it for use in Slack.
  • Build an app exclusively in Slack, entirely outside of Salesforce, employing any number of available AI features.

Agentforce: Agents from Salesforce​

Agentforce allows you to build and customize autonomous AI agents powered by the Salesforce platform via no/low-code with the Agent Builder or coded solutions. Build these with the Agentforce Agent Builder and further customize them using the Salesforce-provided standard Slack actions or code your own custom Slack actions, then deploy them for use in Slack.

AI in Slack apps​

Developing and using some AI features require a paid plan, despite being visible in the app settings on any plan.

Don't have a paid plan? Join the Developer Program and provision a fully-featured sandbox for free.

With the right features enabled, Slack apps can make significant use of AI in their product offerings. Toggling the Agents & AI Apps feature on in the app settings allows you to utilize the entry point and split view surfaces intended for apps using AI in Slack. Code your app to interface with an LLM and have the user-app interaction take place in the split view to have the app work alongside your users in the flow of Slack.

An app that uses platform AI features provides an interface to AI for Slack users. It can communicate with external AI sources, such as OpenAI and Anthropic, as well as access Slack data to source relevant answers to user queries. Apps can be programmed to take actions on behalf of users, whether that is by messaging a channel, sending them a reminder, or even creating a canvas with requested content. The functionality of apps using platform AI features works alongside bot functionality in Slack to streamline tasks.

You can also build in AI access points with app features like modals, shortcuts, slash commands, and events, to name a few. Check out the callout sections titled "Try it with AI" on those pages to see examples using the Slack Bolt framework.

Learn how to build an app with AI features in Slack with the Developing apps with AI features guide.

AI features for Slack apps​

Enabling the Agents & AI apps feature in your app's settings gives your app a new entry point, a split view surface, loading states, suggested prompts, and app threads.

Additionally, the Messages tab typically in the app home is replaced with Chat and History. Past conversations only show in History if the user has sent a message in the thread (threads with only the app's initial message do not show in History).

App entry point: Users can find and open apps through a new entry point in Slack.

Image of an entry point

Split view: Users can initiate private conversations with apps using AI features in a split view within a channel, in the flow of work.

Image of split view

Loading states: Developers can set loading states while application operations are in progress.

Image of loading states

Suggested prompts: Developers can define default suggested prompts to help start the conversation.

Image of suggested prompts

App threads: These are designed to organize conversations and feature a single thread in the conversational experience. Slack automatically starts new threads when appropriate.

Image of ai app threads

Text streaming: Stream lengthy LLM messages to provide a user experience typical of LLM tools.

Three Web API methods work together to provide users a text streaming experience:

If you're a Python or JavaScript fan, our Bolt frameworks and SDKs in those languages have a streamer utility to allow you to quickly implement the functionality of these API methods into your apps.

Gif of text streaming

Block Kit blocks: Use context actions blocks, icon button blocks, and feedback button blocks alongside AI interactions to provide your users the option to give feedback on the AI responses.

Image of context block

Next steps​

➡️ Learn how to set up and manage Agentforce in Slack, including how to connect Salesforce and Slack and add an agent to Slack with this help article.

➡️ Get started building with AI feature offerings in the Developing apps with AI features guide.

➡️ Learn how to further customize your Agentforce agent with Slack actions in this guide.

➡️ Browse which third-party agents are available for Slack in the Slack Marketplace.

➡️ Explore Agentforce documentation.