How Agentforce in Slack Helps Salesforce Scale IT Support with Instant Answers

Executive summary: Salesforce’s Techforce Agent, an AI-powered IT support solution embedded in Slack, is revolutionizing internal IT by providing instant answers and automating routine tasks for over 28,000 employees. Techforce Agent resolves 35% of IT issues instantly, saving 16,800 hours annually and enabling the team to focus on more complex, high-impact work.

The best IT support is prompt support. When your system crashes right before a meeting or you can’t access a critical file, you don’t have time to wait. You need solutions now. But delivering that kind of quick, seamless, and reactive support at scale takes serious coordination and a highly efficient team ready to jump in.

At Salesforce, our internal IT support team, Techforce, fields more than 25,000 requests every month. Until recently, every request required a human agent to respond, no matter how simple the issue.

The Techforce team realized that too much time was being spent solving the same problems again and again. Employees were waiting on answers they could’ve found themselves. Support reps were stuck fielding repetitive tickets instead of focusing on higher-priority issues. It wasn’t sustainable.

That’s why Salesforce built the Techforce Agent through Agentforce and embedded it directly in Slack, where our employees already go for support. The agent helps employees get instant answers to common IT issues, automates routine tasks, and seamlessly escalates complex problems to a human when needed.

Today, the Techforce Agent is helping more than 28,000 employees resolve issues faster while giving our support team valuable time back to focus on high-impact work.

Simplifying support, one conversation at a time

Before launching the agent, the Techforce team had already made big strides by bringing its support model into Slack. But employees still struggled to get the right answers quickly. Searching for help often meant clicking through outdated help center articles.

With the agent in place, that friction is gone. Employees simply describe their issue in plain language and the agent responds with personalized support, often linking directly to tools like the self-service password reset portal or providing step-by-step guidance.

Since launch, the Techforce Agent has handled more than 64,000 conversations and is now used by 28,000 employees across Salesforce. That’s 37% of the company, and growing. The agent currently resolves 35% of issues without human intervention, saving 16,800 hours annually.

“The real power of the agent is that it gives people the exact piece of information they need without the back-and-forth,” said Neil Cummins, Director of Techforce Support at Salesforce.

Not just built into Slack—built for Slack

Behind the scenes, the Techforce Agent runs on Service Cloud and uses Slack’s “Bring Your Own Channel” capabilities. That setup allows the Techforce team to keep its full support tech stack behind the scenes, while giving employees a seamless Slack-native experience.

From a user perspective, it’s simple. Type a question in the Slack channel. Get an instant answer. And if needed, escalate to a human with full conversation context preserved, meaning employees never have to repeat themselves.

The result is a smoother, faster experience for everyone involved.

Smarter knowledge, better outcomes

The Techforce team also took a fresh approach to how it manages the IT knowledge that powers the agent’s answers. The team started with more than 2,000 internal articles but quickly realized that too much information was hurting the agent’s ability to find the right answer.

So the Techforce team narrowed the scope, shrinking the knowledge base down to just 600 curated articles written and formatted specifically for AI comprehension.

“When someone says, ‘I need a password reset,’ they usually don’t say the name of the system or tool,” said Cummins. “We had to make sure the agent could infer what they meant and give the right answer, every time.”

This tighter knowledge model also allows the team to monitor trends and continuously improve the agent’s coverage. And with future integrations planned, like incorporating knowledge from Slack conversations, it’s only getting smarter over time.

Coming soon: More automation and broader reach

The current version of the Techforce Agent focuses on surfacing knowledge and guiding users to self-service solutions. But the next version, already in the works, will go further by taking direct action on behalf of employees.

Learn more about how Salesforce is using Agentforce in Slack at scale.