Building the Future of Advertising: A Conversation with Alkimi's Founders

Main Takeaways

• Digital advertising is one of the most demanding enterprise workloads imaginable, and Alkimi is solving its core problems with the Sui Stack.
• Alkimi co-founders Ben Putley and Chandru Narayanan joined Mysten Labs's Roberto Beltramini and Abhinav Garg for a livestream on how Alkimi uses Walrus, Sui, Seal, and Nautilus together.
• Alkimi's infrastructure is already equipped for agentic media buying, where AI agents negotiate and settle deals in real time.

Overview

Online advertising is a marketplace most people never think about, but it runs continuously underneath nearly everything on the internet. 

Every time a page loads, an auction happens. Advertisers bid to show you something. Publishers, the websites, apps, and streaming platforms you use, earn revenue from the results. In between sits a chain of intermediaries, each taking a fee.

The system works, but barely. Advertisers can't always verify where their money went, publishers wait weeks to get paid, and reconciling a campaign typically means finance teams manually aligning spreadsheets across parties that don't trust each other.

Alkimi was founded to fix that. 

Ben Putley and Chandru Narayanan built careers inside this industry before starting the company. They knew the problems firsthand. When they went looking for infrastructure capable of addressing them at scale, they found that Sui was the only realistic fit.

Alkimi joined Roberto Beltramini and Abhinav Garg from Mysten Labs, the original contributors to the network, for a livestream covering how the Sui Stack maps to the specific demands of enterprise advertising, and why that combination is setting Alkimi up for what comes next.

Why Advertising Is Such a Hard Problem

Abhinav framed digital advertising around three compounding challenges. 

The first is volume. Alkimi processes 15 to 25 million ad impressions per day, and each one generates data — campaign details, targeting signals, fraud indicators — that have to be tracked, stored, and reconciled.

The second is that this data is sensitive and large at the same time. Every impression carries information that is both commercially valuable and legally regulated. In Europe, GDPR governs how personal data can be stored and for how long. The data has to be auditable, but not exposed.

“The third is the meta trust problem: Advertisers don't trust publishers. Publishers don't trust advertising exchanges. And agencies don't trust measurement vendors," Abhinav explained. The result is weeks of back-and-forth trying to agree on what actually happened.

Ben added that verification has become its own revenue stream for a reason. Third parties charge to validate what should already be verifiable. "Verification is a feature of a blockchain that you don't have to pay for," he said.

Built for This

Chandru described the Sui Stack not as a blockchain with a few useful features, but as composable building blocks. 

"It wasn’t just, 'Oh, we've got a blockchain, and we're gonna post our transactions there.' The onchain component is just one aspect of an enterprise solution," said Chandru.

What made Sui the right choice is that each piece of the stack addresses a specific problem Alkimi actually has: 

• Walrus for storing transaction data at scale
• Seal for GDPR-compliant access control
• Nautilus for provable offchain computation
• zkLogin for tying identity to wallets without friction 

Together, they gave Alkimi what no single layer could: private execution, verifiable outcomes, and scalable data storage in one integrated system. 

"It provided us with the Lego blocks that we could piece together in order to build an enterprise product," Chandru said. 

For a full breakdown of how each component works, read the Alkimi case study.

A Template for Enterprise Workloads

Abhinav's observation was that Alkimi isn't just a case study in advertising. It's a signal about enterprise workloads more broadly.

The signals worth watching for: a reconciliation problem where teams spend weeks aligning records, large data volumes that need to be stored and proven efficiently, and a multi-party ecosystem where no one fully trusts anyone else's numbers. 

"Blockchain works when multiple parties benefit from sharing a source of truth," he said. Advertising is one of the hardest versions of that problem.

What Comes Next

The most forward-looking part of the conversation was about AI agents. The premise: agents will increasingly negotiate and execute media buys autonomously. 

That creates a problem if the underlying infrastructure can't keep up. If agents operate off siloed databases, there's no shared record of what happened. You've automated the deal and multiplied the reconciliation. As Ben put it: "You may have saved some time on the front in negotiating the deal, but you've added weeks to the end of it."

A shared ledger changes that. Every action is recorded and Sui's object model lets agents move data and value across the same rails simultaneously, rather than as separate steps reconciled later.

Alkimi is also developing an AdFi liquidity pool that would let publishers access payments earlier than the standard settlement window by drawing on onchain liquidity. The ad market, currently around $750 billion, would become a yield source accessible through DeFi mechanics. As Roberto put it: "You're farming the internet."

The Infrastructure Was Ready Before the Market Was

Alkimi works with brands including Coca-Cola, Samsung, TikTok, Visa, and Kraken, and with publishers ranging from Tubi and Paramount to news publications and blockchain-native media.

Chandru put it well: they didn't set out to build an agentic advertising marketplace. They set out to solve specific, painful problems in the industry they knew. The Sui Stack gave them the tools to do that, and it turned out those same tools are exactly what an agentic commerce future requires. The infrastructure was ready before anyone knew to ask for it. 

Watch the full livestream dedicated to Alkimi and Sui here.