AS ADOPTION of artificial intelligence (AI) accelerates globally, Philippine CEOs see AI as an essential fulcrum for growth and aim to integrate it into key business processes. However, while AI offers significant opportunities, large language models (LLMs) alone are not enough for mission-critical workflows because of inconsistent outputs and limited governance.
As organizations of any size become agentic enterprises — integrating humans, agents, apps and data — Salesforce offers a unified platform that unlocks growth and innovation. One of the key issues is translating such a leap in technology development into sales and revenue streams, among other corporate success metrics.
In an exclusive press briefing, Srini Tallapragada, president and chief engineering and Customer Success Officer at Salesforce, said, “We are in the midst of a big technology transformation, and things like the generative AI revolution are among the biggest we have seen since ChatGPT was launched.
“Every time I talk to CEOs — not just here in the Philippines but globally as well — customers say the LLM models are doing very well. They are beating all the benchmarks. The thing is, CEOs are just not seeing their impact on business value, on P&L, for instance.
“Now, it’s okay to have an LLM, but if you don’t have all the systems surrounding it, you may not reap the success you envisioned at the start. You see, for these models to work well, you need context. You have to understand all of these models are trained on public data. They don’t have access to the company’s data because they are trained on publicly available data.
“That’s what we call ‘the last mile challenge.’ It’s filling the gap between what the agent technology is supposed to do and the real business value it delivers.”
Tallapragada further explained how to overcome the “last mile” challenge. He said, “LLMs are a very good technology. It’s raw intelligence, and raw intelligence needs a place to make it real. And that’s what our Salesforce platform called AgentForce does.
“AgentForce takes the raw intelligence of the LLMs and puts it to work in an enterprise context. That’s been our vision and our execution. And it’s been our fastest-growing product.”
Salesforce’s AgentForce bridges this “last mile,” turning LLMs into reliable, context-aware agents that enterprises can trust with complex, high-stakes operations.
“AgentForce is a solution because we built all that infrastructure. It is proven both by our revenue and by our customers in terms of revenue numbers as well as our customers’ success. That’s the big part.
“It also helps answer complicated questions for customers. But more importantly, we are proud of all the thousands of customers who are going live, who are already live and using it. So, to summarize, there is an enterprise gap which, with AgentForce, customers are finally getting business value.”
Abraham Cuevas, regional vice president and country manager of Salesforce Philippines, said Salesforce is growing in the Philippines. “It’s one of our fastest-growing regions. So, we see this as a really key market for us. We see a lot of room for growth. We have very big customers like BPI, PLDT and others supporting us.
“We’ve opened a new office in November, but that’s not the first time that we’ve been engaging in the Philippines. We have been in the Philippines for years.
“Now with our new office, we’ve grown our teams, and we’ve been investing in all aspects of our business, from executive solutions engineering to professional services.
“We are also supporting Tagalog along with English. And we are all from this region, and we’re very proud of our team here in the Philippines.”
Cuevas also clarified that Salesforce targets regulated industries like airlines and telecommunications. “We try to be very fair. Our target market is the range of enterprises in large, medium and small-scale industries. In the Philippines, we have mid-market and we also have SMEs, all across the market spectrum.”
In that sense, the concept of the “last mile challenge” is attuned to the particular business scale of the enterprise.
Tallapragada said, “We fully understand that for an SME, the problems are different because, for one, they have a lot of legacy systems. For an SME, we try to do most of the out-of-the-box thinking for them. And we know the complexity comes when you’re a small enterprise.
“But when those problems happen, whether you’re a small company CEO or a big company CEO, you still want to grow. You still want to scale.
“We take all the technology transformations over the last few years — whether it’s from cloud to mobile to social media to AI — to help any type of customer get value.
“We say, ‘Let’s handle the technology.’ For example, we talk to a sales rep or head of sales and say, ‘This is how AgentForce can help you generate more leads, help convert more opportunities.’
“We go to the head of support and say, ‘This is how you can provide better proactive support so you can cover more people in less time.’”
“So, once we talk business language, which they understand — like business value — customers come around and agree, ‘I can understand how you can help me.’ And to do that, we need all the technology around, not just the LLM.”
Given that AI can perform 24-hour repetitive work, it could technically outperform human agents and absorb aspects of their skills or intelligence. However, Tallapragada said human workers will remain essential.
“Our mission is seeing humans and agents work together. Without humans in the middle, it’s not happening. Without humans and agents working together, I don’t think we’ll get the right business value. So, I feel strongly that humans plus agents together are required to generate business value.
“Now, will it happen in 20 years? Who knows. But the reality right now is there’s a lot of handholding. In reality, these are very imperfect systems. They’re just very good technology.”