{"id":67411,"date":"2025-06-05T09:55:25","date_gmt":"2025-06-05T16:55:25","guid":{"rendered":"https:\/\/www.salesforce.com\/?p=67411"},"modified":"2025-06-10T12:10:10","modified_gmt":"2025-06-10T02:10:10","slug":"launch-an-agentic-ai-pilot","status":"publish","type":"post","link":"https:\/\/www.salesforce.com\/au\/blog\/launch-an-agentic-ai-pilot\/","title":{"rendered":"How To Launch an Agentic AI Pilot in 10 Steps"},"content":{"rendered":"\n

Launching a pilot is a great way to explore AI agents in a controlled setting. You can create one, test it, and address any challenges before rolling it out. If you want the pilot to succeed, though, you need to do it right: By some estimates, more than 80% of AI projects fail<\/a> \u2014 twice the rate of failure for tech projects that don\u2019t involve artificial intelligence (AI)<\/a>. (Our recent article<\/a> on why many agentic AI<\/a> pilot programs fail offers tips on righting the ship.)<\/p>\n\n\n\n

The best pilots are built on a solid plan that lays out how they\u2019ll be designed, executed, and scaled. But you don\u2019t need to figure this out on your own. Check out our 10-step plan below, as well as guidance and insights from companies that have already done the work, including Big Brothers Big Sisters of America, Engine<\/a>, and Salesforce.<\/p>\n\n\n\n

\n\t
\n\n\t\t
\n\t\t\t

Agentforce World Tour Melbourne is coming<\/h2>\n\t\t\t\t\t\t\t

The agentic era is here, and now is the time to power up with AI agents. Explore expert-led sessions, demos and hands-on training, and find out why businesses are rushing to deploy Agentforce. All in one day, and all for free.<\/p>\n\t\t\t\n\t\t\t\n\t\t\t\t\t\t\t

\n\t\t\t\t\tRegister now<\/a>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\n\t\t
\n\t\t\t\t\t<\/div>\n\t<\/div>\n\n\t\t\t
<\/div>\n\t\n\t\t\t\n\t\t\"\"\n\n\t\t\n\t\t\t\t\t\"\"\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\t\t\t\"\"\n\t\t\n\t<\/div>\n\n\n\n

1. Find the right use case<\/h2>\n\n\n\n

If there\u2019s one piece of advice Travis Gibson, chief technology officer at Big Brothers Big Sisters of America (BBBSA), has for companies that want to launch a pilot, it\u2019s this: \u201cStart with the problem,\u201d he said, \u201cnot the technology.\u201d <\/p>\n\n\n\n

BBBSA easily identified a process it found challenging. For more than 100 years, the nonprofit\u2019s mission has been to match adults (\u201cBigs\u201d) with young people (\u201cLittles\u201d) in a mentoring relationship that benefits both<\/a>. About 135,000 kids currently have a Big, but BBBSA has a waiting list of 30,000 Littles. <\/p>\n\n\n\n

In the past, matching specialists needed six to eight weeks to review all the information they\u2019d collected about Bigs and Littles, including their common interests and backgrounds, and make a match. Gibson and his team realised that an AI agent could do this more quickly. So, BBBSA used Agentforce<\/a>, the Salesforce platform for building AI agents, to create one. \u00a0<\/p>\n\n\n\n

The team built its agent last November and offered it as a pilot to 15 agencies in March. The agent analyses the data and suggests 8 to 10 potential matches for each kid, explaining why each match might work. Human specialists then review the recommendations and make the final decision.\u00a0\u00a0\u00a0<\/p>\n\n\n\n

2. Ask whether agentic AI is the right choice<\/h2>\n\n\n\n

Once you\u2019ve identified the problem you want to solve, ask yourself, which is the best choice: generative AI, predictive AI, a chatbot, or an agent? <\/p>\n\n\n\n

Generative AI is great for creating content, such as reports or emails, while predictive AI excels at using data and algorithms to predict outcomes and events. Chatbots can handle preprogrammed tasks without many variables, such as a simple product return. \u201cBut they\u2019re pretty deterministic, and they can\u2019t behave any other way unless you change the code,\u201d said Irina Gutman, regional vice president, Agentforce accelerator team, Salesforce professional services.

If you need an AI that can make its own decisions, go with an AI agent. It can analyse data, use reasoning, and take action, all while communicating in natural language. \u201cBut there\u2019s inherent variability in agentic technology,\u201d said Gutman, \u201cwhich means you\u2019ll also need to monitor it.\u201d\u00a0<\/p>\n\n\n\n

3. Start small, with a repeatable task<\/h2>\n\n\n\n

\u201cCustomers will say to me, \u2018Give me an example of the sexiest, most complex, advanced agent you\u2019ve ever built,\u2019\u201d said<\/a> Gutman. \u201cAnd my answer is, \u2018Sure, we can do that. But actually, what you want is to start with the most boring, most repeatable, and most low-hanging fruit.\u2019\u201d <\/p>\n\n\n\n

In other words, keep it simple, practical, and impactful. That\u2019s what Engine, a business travel management platform, did with its Agentforce pilot. Engine is known for its stellar customer service, offering 24\/7, no-hold chat and calls. Every year, the company receives 550,000 customer inquiries, 60% of which are initiated via its mobile chat function. <\/p>\n\n\n\n

Engine had grown rapidly and needed to scale, but it wasn\u2019t sure it wanted to add to its head count. To take the strain off its existing customer service agents, the company looked at the easy-to-manage, repeatable tasks it could hand off to an agent. <\/p>\n\n\n\n

\u201cWe had conversations in which we were asking, \u2018Where\u2019s the repetitive work?\u2019\u201d said Joshua Stern, Engine\u2019s director of GTM systems. \u201c\u2018Where\u2019s the work that you can program or teach an agent to handle, as opposed to the work that\u2019s customer-facing, and where it makes the most sense for our human agents to be spending their time?\u2019\u201d<\/p>\n\n\n\n

The answer: An agent could address customers\u2019 most common request, \u201cCancel my reservation.\u201d It\u2019s a narrowly defined, repeatable task that\u2019s perfect for a pilot, and it requires the autonomy and decision-making skills of an AI agent. For example, the agent would need to decide whether a customer was eligible to cancel a reservation. <\/p>\n\n\n\n\n