Keep pulling the thread on Brian Gavin.
Slackbot has evolved from a simple notification tool into a sophisticated, agentic AI assistant. It can now understand complex, natural language prompts, access the full history of a user's conversations (respecting permissions), and execute tasks in other enterprise systems, functioning as a true digital colleague.
The new AI capabilities are delivering massive, quantifiable time savings, with some users reclaiming 20 hours per week. This is shifting the role of the employee from a 'doer' of functional tasks to an 'orchestrator' of human and AI agents, focusing on higher-value strategic work.
Slack is positioning itself as the central "system of engagement" for the agentic enterprise. By acting as an MCP (Multi-agent Controller Protocol) client, Slackbot can connect to and orchestrate various specialized agents, whether built on Salesforce's Agentforce platform or from third parties like OpenAI, Anthropic, and Perplexity.
The platform leverages the vast repository of unstructured conversational data as the 'long-term memory' of a company. The new Slackbot can tap into this historical context to provide nuanced, informed answers and perform tasks without requiring redundant explanations from the user.