Lazy Basically, there is a new appearance on how artificial intelligence agents access and use enterprise data, which launchs the new platform capabilities, which allow developers to work directly to the workplace channels that are directly discussed by the workplace channels.
The company announced Wednesday that its new Real Time Search API And Model Context Protocol Server Third -party developers will provide a safe, secure part of the workplace, conversation, messages and files of the slack of the files, to be safe, permitted. The move has assumed that the discussion data – informal conversations, decisions and institutional knowledge that accumulate in the workplace chat – will become fuel that makes AI agents really useful rather than normal.
"Agents need more data and real compatibility in their answers and actions, and it is coming from the context, and it comes from the context, clearly, within an enterprise conversation," Rob Simon, Chief Product Officer of Slack, said in an exclusive interview with Venture Bat. "And the best place for these conversations inside the enterprise is slow."
With mixed results, this announcement reaches as an enterprise software companies, running to embed the AI’s capabilities in your platform. While the tools like Microsoft’s Coopelot And Google’s gym An important resonance has created, AI agents have been obstructed by adoption that most teams provide a general response from a specific work context.
Slack’s view represents a different philosophy: Instead of building AI characteristics in isolation, the company is positioning itself as the primary layer where AI agents can access non -imposed conversation, which contains the original context of modern organizations’ decision -making.
How does Slack AI plan to unlock the conversation data at the workplace for agents
Technical abilities Lazy Unveil the company as a fundamental problem facing thousands of companies building AI agents: how they should remember to use employees instead of standstone tools instead of making useful in the original flow of work.
Real Time Search API AI applications list the properties of information -related information related to authentic users to inquire from slack data, search for messages, channels, files, and silica canvas and real time. Unlike traditional API, developers need to sew multiple ends simultaneously, the new system provides a single, concentrated method of recovering keywords or natural language indicators.
"It avoids the need to duplicate the slack data between the system, which enables features such as fed up the search," Simon explained. "Therefore, this is a very focused, based on the issue of use that puts the data resident in the slack with appropriate permission and provides access to it as per demand."
Model Context Protocol ServerMade on an open quality manufactured by anthropic, it standards how large language models and AI agents discover and process tasks within the silic, when complicated developers face complexity when building integration into several enterprise systems.
Well -known AI companies are already developing these capabilities. Anthropic’s cloud can now find silicwork spaces to provide a familiar response to the team’s original conversation. Google Agent Space Platform Uses Real Time Search API To produce smooth information between Slack and Google’s AI agents. The enterprise of anxiety now is the basis for its web search capabilities in the team’s discussions, while the dropbox dash provides real -time insights in both platforms.
Why Enterprise Security concerns cannot be removed from Slack’s AI ambitions
The platform’s security architecture indicates what can be a major concern for enterprise users: AI agents just accessing information that are authorized to view users. Slack’s point of view depends on the authentic access, which respects the current permitted structure.
"The basic method is that information is accessed by the user," Semon said. "When one of these agents calls the call back to the slack, the user confirms the agent, which then verifies the slack using the user’s credentials."
This means that AI agents can only access messages, private channels and public channels that are already allowed to view the certified user. In addition, Slack has contracted on the use of API response to AI models training, and concerns about sensitive enterprise data are being used to improve the third -party AI system.
The security model is particularly important in view of the central status of the silic in the enterprise workflower. The platform has become an operational backbone for countless organizations, which produces a vast reservoir of sensitive information, which includes strategic decisions, confidential debates, and institutional knowledge that require cautious access control.
For international consumers, maintains slack Data residence In many areas, to act locally information to meet the needs of autonomy, sovereignty. Of the company Enterprise Plus The project includes the features of comprehensive security and compliance designed for regulated industries.
Microsoft teams face new pressure as Slack AI accepts ecosystem strategy
With this announcement has been represented by the latest move of Silk in a rapidly intense competition Microsoft teamsWhich is aggressively incorporating AI’s capabilities through its Co -Pilot Platform. While the two companies are embedded AI in their support platforms, they are clearly taking different perspectives.
When asked about competitive dynamics, Simon emphasized the user experience compared to the feature comparison: "People like to use the slack. So they like its original last user experience. They also like to experience their other software in the slack, and therefore people like the approval of the spending reports in the silic, and they like to approve travel requests and make pig tickets, all of which is in the flow of work."
Instead of creating a comprehensive suit of productive tools such as Microsoft, Slack’s strategy focuses on becoming the focus of integration, where other software experiences are combined with each other. This approach has already shown the results, with the company noted that the Agent Startups have obtained "10s of the 1000s who have installed it in 120 days or less" Building in the silk market.
Time also reflects the wider dynamics of the market. Sales force, which Gained silic in 2021 .7 for 27.7 billion, the platform is presenting the central position of its AI strategy, while its product portfolio increases prices. In June, the company increased Slack Business+ Pricing Each month from 50 12.50 to $ 15, the second price increase in the age of 24 months.
Slack’s amazing income strategy: No fees for AI developers
Unlike some platform companies that take shares from third -party developers, Slack has not chosen partners to make money from their AI capabilities directly through fees. Instead, the company’s revenue model focuses on deep consumer engagement and maintaining.
"We do not do tax sharing models with our partners," Semon said. "The advantage of the slack is that people use their software maximum use within the slack, and users are busy on our platform. We want them to have a great experience in doing their job in the silic."
This approach reflects a wider strategic calculation: by making the slack the most attractive platform for AI development, the company can increase its price as the central nervous system of enterprise work, which can justify advanced prices and reduce customer.
The strategy is working. Slack reports that more than 1.7 million apps are actively used on its platform every week, 95 % of users have said that the use of the app in the silk makes these tools more valuable.
What could be the meaning of AI discussed for the production capacity of the enterprise
This announcement shows how the capabilities of the Enterprise AI will be deployed and tested. Employees learn how to use separate AI tools for various tasks, with Slack’s vision accessible through the same interface used for human cooperation as teammates who communicate AI agents.
"You can imagine a time where we will all have a series of agents working by us," Semon said. "They need to stop you. You have to intervene and change what they are doing in reality – may be fully redirected. And we think Slack is a great place to do this."
The dialogue approach to the AI ​​talks can solve one of the biggest challenges to adopt enterprise AI: Switching costs in context that reduce production capacity when employees need to move between several special AI tools. By maintaining AI’s conversation within the current communication flu, the Slack aims to reduce the academic overhead of working with several AI agents.
The focus on the platform interaction data also solves a significant limit to the current enterprise AI system. Although many AI tools can access structural data from database and enterprise software, informal conversations where real decisions are made and institutional knowledge are shared has been largely accessible to AI systems.
Behind the curtains: How Slack created infrastructure for real -time AI questions
Behind the scenes, Slack has created a technical infrastructure designed to handle the requirements of real -time AI questions while maintaining performance for its basic messaging capabilities. This system includes rate limits for API calls and restrictions on data volume that can be returned in response to questions, ensuring that the search is intensified and targeted rather than trying to take action on the entire dates.
"When someone searches on real -time search API, we will not return the entire Silk Corps," Simon explained. "It will be highly targeted, related to the ranking and this particular question. We are doing this so that we can basically guarantee the time for a fast -paced response."
The developers, the Setup process, is upright, which requires only one validation and app to create the current silicing integration. Low obstruction in admission can accelerate the adoption of AI startups and enterprise software companies’ growing ecosystem that is seeking to embed the interactions AI capabilities.
The success of the expansion of Slack’s AI platform will depend on whether businesses have accepted the AI ​​discussing as a natural expansion of team communication, or they prefer more made methods offered by rivals. Since the enterprise software companies continue to racing AI’s capabilities, the company that best solves adoption and context issues can be exposed as the basis of AI -powered work.
But for now, the Slack has made his choice clear: In the AI’s supremacy, the winner will not be determined by the very sophisticated algorithm – whatever control it controls.