Zip AI Summit: Reshaping procurement

Unlocking the future of autonomous procurement.

For Zip, the mission couldn’t be clearer.

The AI platform for enterprise procurement has scaled significantly over the past few years and announced the launch of three new innovations at its AI Summit in New York City in June.

Through keynotes, case studies and workshops focused on AI-driven automation, Zip explored how organisations can unlock new efficiencies in procurement. The half-day event at The Ace Hotel in Brooklyn also featured networking opportunities and live demonstrations of Zip’s AI agents deployed by leading customers, such as OpenAI and Datadog.

Zip believes that intelligence alone is not enough for enterprise procurement. AI needs governance, context, auditability and orchestration to undertake lasting and scalable work inside financial processes. This is a theme which underpins Zip’s newest launches, Zip Superagents, Zip MCP and AI Spend Automation. 

Designed to work alongside procurement teams, Superagents are an AI platform that takes on real work – reviewing contracts, unblocking approvals, coding invoices, and researching vendors – with every action audited and governed inside Zip. Throughout the day, attendees were shown how these tools really work in practice. For example, the Intake Superagent guided an employee through a request in natural language, answering policy questions and gathering what approvers needed. The Contract Superagent reviewed agreements, flagged risks and recommended redlines and the Procurement Superagent unblocked stalled requests and managed supplier communications.

Next, Zip MCP is a procurement-native implementation of the Model Context Protocol that connects Zip’s procurement intelligence directly to AI assistants like Claude and ChatGPT, enabling employees to work in their preferred AI tools without leaving Zip’s controls behind. Lastly, Zip also unveiled its AI Spend Automation, a new enterprise offering that combines platform access, AI consumption credits, and support from Zip’s own forward-deployed engineers to help the most progressive organisations fully automate their intake-to-pay process.

Speaking exclusively to CPOstrategy at the event, Felix Meng, Co-Founder at Zip, reveals that over the next five years he expects procurement teams to be leaner than ever before and will deliver 10X of today’s impact with fewer resources.

“At Zip, we believe in the next few years, 90% of the millions and millions of requests and approvals happening in our system every day will be at bare minimum AI-assisted, if not fully automated in those steps,” he tells us. “What that unlocks for the procurement organisation, stakeholders and clients internally is less and less time is going to be focused on a lot of these manual triage type tasks. That’s really at the beginning of the intake and request initiation and at the tail end when you’re actually getting into procure-to-pay and payment of the vendor. 

“More and more time is actually going to be focused on the middle where there’s a lot of relationship management depth, relationships with stakeholders, with internal requisitioners and leaders as well as relationships with suppliers. I don’t think those are the areas where AI will be heavily replaced. It’s just going to ultimately open up a lot more time for your average practitioner to spend time on the stuff that they enjoy.”

Across the day, CPOstrategy spoke with a number of Zip leaders and customers about the new capabilities of Zip’s AI platform and how AI is reshaping the work of procurement.

• What’s your key take-away from this year’s Zip AI Summit? 

Kat Devlin, Head of Procurement Operations, OpenAI 

“Regardless of the size or tech stack of the company, the problems are the same. Procurement teams want it to be easier to serve their customers and the way agents add value is by relieving that tactical burden, so they can show up in a different way.”

Natasha Berk, AI Principal Product Manager, Zip 

“Everyone is still figuring out how to leverage this technology in a way that truly automates their work. Also, without adding additional complexity for audits and doing it in a way that’s governable, which is so important for procurement. I’m really excited to see how agents can help people do that, save time and focus on the parts of the job they enjoy and that add strategic value.” 

Michael Denari, General Manager of AI, Zip 

“Speaking to people here, our focus on agents and their governance, so that we can move from assisting procurement to really automating it, is really resonating. I think a lot of folks are stuck in admin tasks, so we’re really excited to see what customers are going to be able to unlock using our new Superagents platform.”

Jing Jian, AI Design Architect, Zip 

“AI is such a different journey for every company, but procurement has so many opportunities to use AI to make teams more strategic. What excites me is seeing what our customers build on top of our Superagents platform. That’s what really keeps me excited.”

Felix Meng, Co-Founder, Zip 

“We had a number of panel speakers today from forward leaning companies across all industries, and their message was that they’re just getting started. You might think they’re so deep into this, it’ll be hard to catch up, but we’re all just getting started. So, it’s not too late to start and it’s certainly the right time to, because there’s so much educational momentum.”

• What processes are you most eager to automate with AI?  

Lesley Garafola, Managing Director, Global Head of Procurement and Vendor Management, Barings 

“For us it’s the intake. It’s meeting our employees wherever they are. I don’t want procurement to be a bottleneck. It’s allowing Zip to think about what a stakeholder needs, then moving them through that process. Today it takes almost two weeks to get through the intake process, so we’re looking at building that out, as well as our purchase order process in partnership with our finance team.”  

• What makes this generation of Zip agents different?  

Natasha Berk, AI Principal Product Manager, Zip 

“Our first generation of Zip agents executed deterministically and the big difference now is these agents execute dynamically, so they’re able to operate like a human user and adapt to the task at hand. They’re also able to collaborate with other stakeholders in Zip, which is going to be really impactful.”

• What did five years on the customer side teach you that you think most vendors get wrong?  

Michael Denari, General Manager of AI, Zip 

“The AI wave created a very strong push from vendors to release AI capabilities before they really got their foundations right. There are so many Copilots that aren’t well connected to back-end data. So, making sure AI capabilities are focused on driving impact, versus racing to market, has been a big learning.”

• How do you build the kind of flexibility that ensures an agent can handle the unexpected and still have people trust it?  

Jing Jian, AI Design Architect, Zip 

“Flexibility and trust are really interesting levers. They work hand in hand and I think about them in layers. You want flexibility in the agent reasoning layer to make sure the agent is agile enough to take in new inputs and adjust its course of action, but you probably want a lot more trust and governance around the actions it takes. There’s a difference between read and write. I think that’s what makes Zip agents really special, the governance and compliance layers.”

• Why is governance a key factor making AI for procurement successful?  

Sophia Jimenez, Procurement, Anthropic 

“I’m an ex-auditor, so governance is extremely important to me. I don’t think of agents as a replacement for any of the review work, or the work we do in procurement, but more of an assistant that calls our attention to certain things. So, governance is essential to make sure we’re covering all our bases, especially with vendor relationships and accounting.”

• The world’s leading AI companies run their own procurement on Zip. What does it mean that the builders of AI trust it to govern their own AI-assisted workflows?  

Felix Meng, Co-Founder, Zip  

“They need a platform that has a lot of data context. You could build any number of agent automations outside that platform, but the problem is you’re not bringing that to life and meeting employees where they’re conducting their work. Beyond that, they value the safety, governance, and consistency of making sure agents are conducting the exact type of work you want them to, every time. And Zip does that.”