This article is part of VentureBeat’s special issue, “AI at Scale: From Vision to Prosperity.” Read more from the issue here.

Just seven or eight months ago, when a customer called or emailed Reading Systems with a service question, a human agent handling the question starts searching for similar cases in the system and analyzing technical documents.

This process takes about five to seven minutes; then the agent can offer the “first meaningful answer” and finally start troubleshooting.

But now, with AI agents operated by Salesforcethat time is shortened to a minimum of five to 10 seconds.

“That’s a big (decrease),” Andrew Russo, enterprise architect at Baca Systems, told VentureBeat. He emphasized that, “for us, it’s not about how to eliminate the number of people, reduce the staff. Our goal is, how can we make sure that the customer can get back up and running as soon as possible?”

Closing time gaps, delivering faster resolution times

BACA Systems, a robotics manufacturing company based in Michigan, first implemented Salesforce in 2014, eventually adding Cloud Services to replace the “vanilla, or maybe more like strawberry ice cream, basic service cloud,” Russo explained. The company made a “giant digital transformation” in 2021, bringing in Salesforce’s resource planning (ERP) platform.

Team members soon began working with predictive AI for sales and manufacturing forecasts; then the company evolved into AI agents, implementing Salesforce Agentforce within the last year.

An initial use case for the key is service calls. Russo explained that about 57% of the questions that come in from customers are hardware-related (for example, a machine that is down or needs calibration).

Now, instead of sifting through databases for past customer calls and similar cases, human representatives can ask the I have an agent to find relevant information. AI runs in the background and allows humans to respond quickly, Russo said.

AI can also support continuity control. For example, a circuit breaker may be constantly tripping, indicating that there is a short in the wire that needs to be investigated, Russo explained. This will help eliminate ongoing issues that have not been resolved in the past.

“It’s all about how we deliver faster resolution times for customers,” Russo said.

AI agents generate sales leads, handle customer inquiries

Another critical use case is sales, because as a small company, Baca naturally does not have hundreds of sales people or even dozens (in fact they have less than 10).

“We have a boatload of leads that we don’t have time to actually contact,” Russo said. “Our goal is: How do we start making it?”

AI can serve as a sales development representative (SDR) to send general inquiries and emails, have a back-and-forth dialogue, then forward the prospect to a member of the sales team, Russo explained. Bringing in additional salespeople to handle such tasks requires tens of thousands of dollars for salaries, but if AI can create new deals, its cost “is very easy to justify.” .”

In the coming months, the company plans to deploy customer-facing service agents which can interact with human users via text message to open and manage cases without the initial need for human intervention. If the AI ​​agent cannot solve a problem, it escalates the issue to a human representative.

The intent is, “How can we continue to provide more value to customers on the service side and create more deals on the sales side?” Russo announced.

Outside of sales and service, Baca uses AI to generate emails, create receivables and create “very serious collection letters” when needed. Russo, for his part, uses technology for checking the deduplication feature, exploiting retrieval-augmented generation (RAG) with quick builders to find duplicates to prevent bad data from – ported to Salesforce.

There was little or no pushback from employees, he reported: The company started small, initially giving a select group of users access. Others then started asking questions. “They really started asking (us) to give them access,” Russo said. “No one is afraid of it; they love using it because it helps them do their work better.”

The company continues that deliberate, incremental approach as it further incorporates AI to keep it agile. “Our goals haven’t changed, just how we got there and the path we took,” Russo said. “It’s a different road, it’s a better road – it’s the highway.”

AI serves storage for ezCater

Corporate catering is more complicated than it sounds. There may be changes in head counts, dietary preferences and dietary restrictions, as well as other logistical challenges.

ezCater continues to grow in this space, and as it is, the level of high-touch service required to solve customer needs can be difficult to scale without the use of technology, Erin DeCesare, CTO of the platform of workplace catering, told VentureBeat.

But once the company implements Salesforce’s Agentforce, a customer who needs to change an order will be able to communicate their needs to the AI ​​in natural language, and the AI ​​agent will automatically make the changes. . When more complex issues arise – such as a change to an order or a complete change of location – the AI ​​agent can quickly push the matter to a human representative.

“It’s a huge cost savings for us,” DeCesare said.

Another intended use case is “restaurant discovery” – that is, AI agents is able to guide users to the best place based on inputs about their food preferences, budget, location and other factors. It’s backed by data from millions of workplace food orders. “It’s perfect for NLP and AI,” DeCesare said.

ezCater initially engaged in-house AI agents to assist concierge agents, and the human agents wanted it, he reported. “We give them the tools to be better, and be able to handle more calls.”

There is a shift in the comfort level of engineers, too, because they are able to think of agents more structured. “They can try and trust the way it feels like software development,” DeCesare said. “It’s like what they expect in the software development lifecycle.”

Business partners are also excited about the possibilities for tasks such as business analysis or process maps. “Technology has become very accessible in the last six months,” DeCesare said. “You can easily see how this can quickly become the norm. We’ll be in a very different world 12 months from now.”