How teams use Slack AI agents (2026)
Davis Christenhuis
Most teams already use Slack as their primary workspace. But answering the same questions repeatedly, searching for information across tools, and keeping everyone updated takes time.
AI agents extend that workspace by surfacing answers, retrieving context from connected systems, and automating repetitive work. This guide covers what Slack AI agents do, how they work, and how to build one that fits your team.
📌 TL;DR
Looking for the summary? Here are the key takeaways:
- What they are: Slack AI agents respond to questions and automate tasks directly inside Slack channels, pulling information from connected tools and company documentation.
- Two types: Slack AI includes 11+ built-in features (summaries, search, recaps, Slackbot, and more) for Slack data. Agentforce, a separate Salesforce product, also accesses CRM and third-party data.
- How teams use them: Engineering teams get instant code guidance, support teams answer product questions without manual searches, and sales teams access collateral and customer data by @mentioning agents in Slack.
- Proven results: Alan reduced project time by 10-20%, Malt cut ticket closing time in half, and Assembled saved hundreds of hours searching and finding company information by replacing cross-team interruptions with agent queries.
- Quick setup: Connect your data sources, write plain language instructions for your agent, install the Slack integration, and start using it in any channel.
💡 Want agents that know your full tech stack? Start your free Dust trial →
What is a Slack AI agent?
A Slack AI agent is software that responds to questions and performs tasks directly inside Slack channels, DMs, or threads. These agents use natural language processing to understand requests and either retrieve information, generate content, or trigger workflows.
Slack AI agents fall into two categories: built-in AI features that work primarily with your Slack workspace data, and third-party agents that connect to your broader tech stack like Notion, GitHub, Google Drive, and CRM systems.
Agents can summarize long conversation threads, answer questions by pulling from company documentation, draft responses based on context, and automate routine tasks like creating tickets or updating project status.
What you can do with Slack's built-in AI
Slack offers a growing suite of built-in AI features, plus deep integration with Salesforce's Agentforce platform.
Slack AI provides native features for every employee on Business+ and Enterprise+ plans, including channel and thread summaries, AI-powered search, daily recaps, file summaries, message translations, huddle notes, and canvas content generation.
At the center of it is Slackbot, a personal AI agent that can:
- Summarize channel conversations and threads
- Draft updates, build professional briefs, and create content in your tone of voice
- Analyze files uploaded to Slack
- Tailor responses based on your Slack activity and the content you have access to
Agentforce, built by Salesforce, brings customizable AI agents into Slack. These agents are grounded in both Slack conversation data and Salesforce CRM data, and can be extended with third-party tool data through Enterprise Search. Note that Agentforce requires a separate Salesforce license and is configured in Salesforce's Agent Builder.
Pre-built templates are available across departments, including:
- HR
- IT
- Customer service
- Sales
- Marketing
- Legal
- Product & engineering
- Supply chain management
Agentforce agents can create channels, update Slack canvases, and send DMs based on triggers or @mentions.
For teams that want agents connected to the rest of their knowledge stack, here's how to go further.
How teams use Slack AI agents
Dust is an AI platform for building custom agents that connect to your company's knowledge stack. You can call these agents directly from Slack to answer questions, retrieve context, and automate tasks.
Unlike native Slack AI, Dust agents pull from Notion, GitHub, Google Drive, Salesforce, Zendesk, and other tools, so they can answer questions that require context outside Slack.
For engineering teams
Engineering teams use Slack AI agents to answer technical questions without interrupting colleagues or switching between tools.
Common use cases include:
- Surfacing code examples from GitHub
- Retrieving documentation from Notion or Confluence
- Referencing past decisions from Slack threads or project management tools
- Onboarding new developers with instant answers to recurring questions
- Troubleshooting bugs and finding internal best practices
Example: Alan's Slack agent
Alan, a digital health insurance platform, built a Slack agent called @code-help to give engineers instant access to their codebase. The agent pulls from Alan's codebase, internal documentation, and past Slack threads to answer technical questions.
Engineers @mention the agent in their Slack channels to ask technical questions. For example, engineers ask questions like "How do I get the current user ID?", which Vincent Poulain described as his "aha moment" with Dust.
The agent responds with contextual answers aligned with Alan's coding style, best practices, and context from previous Slack conversations.
The result: Alan's engineering team reduced project completion time by 10-20%.
For operations and support teams
Operations and support teams use Slack AI agents to answer internal questions about products, policies, and processes. Instead of forwarding questions to subject matter experts or searching through documentation manually, employees can ask an agent and get instant answers.
Common use cases include:
- Answering product and billing questions from internal teams
- Providing compliance and policy guidance across multiple regions
- Reducing repetitive support tickets from colleagues
- Onboarding new team members with instant access to company knowledge
- Maintaining consistent responses across languages and time zones
Example: Malt's Slack agent for internal support
Malt, Europe's largest freelance marketplace, embedded a Dust agent in Slack to handle internal product questions. Before the agent, employees would ping the support team for answers about contract terms, tax details, or product questions across different countries. The support team was overwhelmed by repetitive requests that could have been resolved with better knowledge accessibility.
Now, employees @mention 'Malty AI' in Slack and get instant answers pulled from Malt's Notion-based knowledge base and public FAQs. The agent handles everything from tax details to product feature explanations.
Malt also built a feedback loop using Slack emoji reactions. When someone uses the agent, they can rate the response with emojis like "Good," "Bad," "Average," or "Out of Scope." The support team reviews feedback to improve the agent's knowledge base.
The result: 100% of Malt's customer experience team now uses Dust agents daily, and the support team cut ticket closing time in half.
For sales and marketing teams
Sales and marketing teams use Slack AI agents to access product information, competitive intelligence, and customer data without interrupting subject matter experts.
Common use cases include:
- Finding the latest product decks and sales collateral
- Retrieving messaging guidance and positioning documents
- Accessing customer research and competitive intelligence
- Answering questions about product features and roadmap updates
- Onboarding new team members without relying on tribal knowledge
Example: Assembled's Slack agent for cross-team knowledge
Assembled, a customer support platform with 95% internal Dust adoption across 120+ employees, used Dust to eliminate constant cross-team interruptions across their teams. Sales reps, customer success managers, and new employees would ping marketing and CS with questions like "Where's the latest product deck?" or "What's our messaging on feature X?"
Assembled uses Dust agents that connect to Google Drive, Notion, Slack, Salesforce, Gong, and other tools, giving every employee a single place to find information across the company's full knowledge stack.
The result: The shift transformed how teams work together. Cassandra Stumer, Product Marketing Manager at Assembled, says: "Just in terms of searching and finding company information alone, Dust has saved us hundreds of hours."
💡 Want to see how other companies use agents with Dust? Check out all customer stories →
How to set up a Dust agent in Slack
Setting up a Dust agent in Slack takes five steps:
- Create a Dust workspace: Sign up at dust.tt and complete workspace setup. You can start with a free trial to test the platform.
- Connect your data sources: Navigate to your Spaces and add connections for the tools your team uses — Notion, GitHub, Google Drive, Zendesk, Salesforce, Slack, or any other platform where company knowledge lives. Dust syncs data automatically so agents always have current information.
- Build and configure your agent: Create a new agent and write instructions that define what the agent does. You can add Slack as a tool so the agent can read and interact with Slack directly. For example: "Answer questions about our codebase using GitHub repositories" or "Help support reps draft responses using our Zendesk knowledge base."
- Connect Slack to your workspace: Your workspace admin enables the Slack integration, and once connected, you can link your agent to specific channels from its setting.
- @mention your agent in Slack: In any channel where the Slack bot has been added, type
@dust ~YOUR_AGENT_NAMEfollowed by your message to call a specific agent.
In this example, the LinkedIn writer agent responds with structured guidance on posting times, sourced from its connected knowledge base.
💡 Ready to build your first Slack AI agent? Try Dust free for 14 days →
Frequently asked questions (FAQs)
Do Slack AI agents work across all my Slack workspaces?
Each Slack workspace requires its own integration setup. If your company uses multiple Slack workspaces, you need to connect the AI agent separately to each one. Once connected, agents can respond in any channel where the bot has been invited. Most platforms let you reuse the same agent configuration across workspaces without rebuilding from scratch.
Do Slack AI agents learn from past conversations?
It depends on the agent. Slack's native Slackbot uses your Slack activity to tailor responses, but it doesn't retain individual conversation history after generating an answer. Third-party agents typically use retrieval-augmented generation, which means they pull information in real time from connected sources rather than learning from past interactions. Some platforms let you save feedback through emoji reactions to improve future responses, but most agents don't store conversation history for training purposes. Your data remains in your workspace and is not used to train external AI models.
Do I need developer skills to build a Dust agent for Slack?
No. Dust agents are built using plain language instructions. You write what the agent should do, connect your data sources through pre-built integrations, and deploy. The platform handles the technical setup. Advanced users can add custom workflows or API calls, but it's not required for standard use cases.