Skip to content
Nancy Xu, Salesforce, VP AI & Agentforce on Autonomous Recruiting artwork

Nancy Xu, Salesforce, VP AI & Agentforce on Autonomous Recruiting

Venture with Grace

April 17, 2026

Nancy Xu is VP of AI & Agentforce at Salesforce after selling Moonhub. ~~~~~~~~~~~~~This episode is brought to you by Nebius — the ultimate cloud for AI innovators.

Speakers Grace Gong, Nancy Xu

TopicsInvestingBusiness

Grace Gong (0:04)

Oh, we're on. Hi Nancy, welcome to Venture with Grace.

Nancy Xu (0:08)

Hi Grace, thanks for having me.

Grace Gong (0:09)

Oh my God, we are currently at the Salesforce Conference. So maybe we could start with your background. So obviously, you are one of the leading voices in AI, and you have sold your own company last year. Well, yeah, last year, it was like very-

Nancy Xu (0:22)

2025

Grace Gong (0:23)

2025, you were doing amazing AI recruiting company. Now, you are BPI Salesforce. Maybe we could talk about your journey into Salesforce, and what are you guys doing now?

Nancy Xu (0:33)

Sure. Yeah, I'm so glad you're able to join us at TDX this year. Grace, this is our biggest developer conference every single year. A bit of my personal background. So I joined Salesforce in, gosh, 2025 It feels like a lot longer. I started a company in the enterprise AI agent space prior to this, worked on it for several years. Really, at the time, commercializing research that I had worked on at Stanford during my PhD in AI agents, and then later bringing it back to the enterprise.

Earlier last year, we had the opportunity after racing around, we were approached by Salesforce about how to position it in other conversations, and we felt very lucky to be at Salesforce.

Grace Gong (1:15)

For sure. I'm so excited to hear everything about Agentforce, since I feel like Agentforce has been such a big key things for Salesforce. Maybe we could start with what is Agentforce for the audience who is first time hearing about it, and then maybe we could talk about the development cycle and observability as well.

Nancy Xu (1:37)

Yeah, love to. Really, how I think about Agentforce here at Salesforce, it's really all-incumbency brand for all things AI and agents, and especially as businesses evolve to become more agentic, this is really the core of what we're really thinking about. How do we scale as the next play for Salesforce?

At Agentforce, we started about two, three years ago now, really focused on building agents for customer use cases, so helping customers build those service agents, so sales agents for their enterprises. Now, of course, Agentforce has rapidly expanded. As you can imagine, businesses are looking for agents, not just for sales and service, but for every use case under the sun and for their own employees as well. So we build agents inside Slack for Slack use cases, as well as those employee use cases, as well as customer-facing agents too. So it is really the whole spectrum, as well as of course the infrastructure that we have underlying, that is critical to enable our customers to build their own agents and customize their own resources.

Grace Gong (2:36)

I would say like, I feel like Salesforce is like having clients from smaller company to the bigger company, and you guys are operating at such a high speed as a company. I'm curious, compared to last year, what would you say is the main difference from the development perspective for agents?

Nancy Xu (2:56)

Yeah. I mean, for so much has changed for us as a business.

This time last year, we had 5,000 customers on Agentforce. We now have over 20,000 customers that are actively deploying agents on Agentforce. That growth has been phenomenal. This is the fastest growing product at Salesforce right now. We're really proud about that. But it's not just that. I think like you said, Grace, the entire landscape for agents and how we build AI is changing.

One thing that I'm personally very excited about, that we're investing a lot in at TDX this year, is how do we enable people to actually build agents in the tools of their choice? I think last year, we were all trying to build agents through these UI interfaces. Every company has their own version of an agent builder. Now, that's changing a lot. Now, if you ask any of your developer friends, software engineering is totally different now with pod code and these agentic harnesses. We think the same will be true for agent development. At TDX this year, my team has worked on a lot of new releases that opens up our CLI and CP servers. We're opening up our APIs so that anyone can build on top of Salesforce using these coding agents, using these products, in order to really build that application layer that we think will be really critical for our customers.

Grace Gong (4:15)

I feel like, like you said, because of the development process has been changing so fast as people started adopting weather like cloud or codex. So basically, I feel like developing agents is completely different as well. So I guess which sector do you feel like is adopting agent as the fastest since Salesforce has such a broader, like broad like touch point with every industry and everything? You know, what is like maybe application layer of things that you find very interesting from your perspective?

14 more minutes of transcript below

Thousands of transcripts fetched by people building searchable podcast archives

Fetch the whole transcript

The demo key returns a sample episode in full, no card needed:

request
curl -H "x-api-key: pt_demo" \
  https://spoken.md/transcripts/1000651996090

Markdown with the speakers named, for your notes, your knowledge base, or anything that makes HTTP calls.

From $0.10 per transcript. No subscription. Credits never expire. Prices exclude VAT, added at checkout for EU customers. Not what you expected? Email us within 14 days with 20 or fewer credits used and we refund the pack in full.

Using your own key:

request
curl -H "x-api-key: YOUR_KEY" \
  https://spoken.md/transcripts/1000761909447