Conversational AI is coming to business communications platform Slack with Slack GPT, a new feature which will have the ability to summarise meetings, write notes and spot patterns in sales data imported directly from parent company Salesforce’s Sales Cloud. Slack has had some generative AI for the past month, including versions of ChatGPT from OpenAI and Anthropic’s Claude, but this is the first native integration throughout the platform’s ecosystem.
Announced as part of the Salesforce World Tour NYC, the company says this is part of an entirely new generation for Slack as a platform. Einstein GPT, the enterprise-friendly chatbot built to work on top of the Salesforce platform, will also be available inside Slack for the first time.
Speaking during a press conference ahead of the announcement, Ali Rayl, Slack senior VP said the new Slack AI platform will allow no-code workflows built directly into Slack using any large language model. There are already 26,000 apps in the ecosystem with 4,000 alone using OpenAI’s GPT-4 API.
No code workflows will allow a user to embed a series of actions from a simple natural language prompt “making it easy for anyone to deploy AI automation”. When integrated with Einstein GPT and wider company data, the company claims it will be a major productivity boost.
Mirroring announcements made for other Salesforce products, Slack’s ecosystem will be large language model independent, allowing companies to integrate any model. This includes open platforms from OpenAI, Anthropic and AI21 Labs, or custom-built LLMs like Dolly from Databricks.
“Generative AI has enormous potential to redefine how work is done and unlock significant business productivity,” said Lidiane Jones, CEO of Slack. “The real power of this technology is when AI can analyse and act on the most valuable data from a company’s most trusted resource — its own internal knowledge.”
Slack GPT: a conversational AI platform
Jones described Slack GPT as “the conversational AI platform of the future” which would allow organisations to tap into trusted customer data and employee knowledge to “work smarter and make smarter decisions faster”.
A survey by Slack of 18,000 desk-based workers found that 90% of those using generative AI are likely to report higher levels of productivity but only 27% of companies are using AI tools to increase efficiency within their organisation.
The company says native integration of AI within productivity platforms is the easiest and safest way to deploy generative AI. “Native capabilities are purpose-built AI features designed to fit seamlessly into Slack’s interface, making it more convenient and easier to get contextual AI assistance,” it said in a statement.
One use case suggested during a presentation was to have Slack GPT create a summary of key action points missed since you last signed in or to create a summary and “next steps” following a huddle.
It could also be used with Einstein GPT to create sales plans or marketing pitches based on discussions within a specific Slack channel that uses real company data.
That “real data” is likely to be the biggest sticking point when convincing companies to switch to using generative AI solutions. Recently Samsung banned employees from using ChatGPT and a Gartner report found concerns among enterprise executives over data security.
Salesforce also announced it is partnering with consultancy Accenture to create an “acceleration hub” to provide organisations with technology and experience to scale the use of generative AI for CRM.
“Generative AI has enormous potential to transform the way companies interact with customers and drive business growth,” said Bill Patterson, EVP and general manager, for C360 applications at Salesforce. “We want to come together to bring the visionary and technical expertise needed to unlock new possibilities for AI- and data-powered CRM, driving a profound transformation in customer interactions.”