# The People Who Monetized Mobile Apps Are Building the Ad Layer for AI Chats > Source: > Published: 2026-07-10 00:06:38+00:00 [Tal Shoham](https://www.linkedin.com/in/tal-shoham-6302ba31/), [Amir Shaked](https://www.linkedin.com/in/amir-shaked-34090862/), and [Nimrod Zuta](https://www.linkedin.com/in/nimrod-zuta-18b5a554/) have raised $27 million for [Velocity](https://velocity.io/), a New York and Israel-based startup trying to turn AI chats into the next major performance advertising market. The seed round was led by [NFX](https://www.nfx.com/) and [Red Dot Capital Partners](https://red-dot.capital/), with participation from [Stardom Ventures](https://www.stardom.vc/), [Corner Ventures](https://corner.xyz/), and [Transcend](https://www.transcend.fund/). The investor lineup is a clear bet on operator pedigree: Velocity's founding team comes out of ironSource and Unity, where its founders spent more than a decade building advertising, monetization, mediation, and growth systems for mobile apps. [Guy Kirschbaum](https://www.linkedin.com/in/guykirschbaum/) joined as founding vice president of engineering. Velocity is applying the mobile app monetization playbook to a different interface: the prompt box. The company's pitch is that AI products increasingly resemble search in the way users express intent, while their economics look worse than mobile utilities because every serious interaction carries model serving cost. Subscriptions help only when users convert, and many consumer AI apps are already using usage caps, paywalls, and premium tiers to manage the cost of free users. Shoham is making the comparison explicit. In an interview with CTech, he said Velocity wants to build a monetization infrastructure layer for AI companies similar to the role ironSource played in mobile apps. He also said AI companies face a larger problem than game developers because free usage is expensive to serve. CTech quoted him saying that advertising gives AI companies another revenue channel beyond subscriptions while allowing users to keep using the products. ### What Velocity is actually selling Velocity's product has three main pieces: an AI-native ad network, a mediation and auction layer, and what it calls a conversation manager. On its [homepage](https://velocity.io/), Velocity says its system turns each user message into a creative brief, analyzes the user's intent, matches the prompt to an advertiser, generates a tailored ad, and places that ad natively inside the conversation. The [publisher product page](https://velocity.io/ai-publishers) shows the kinds of inventory Velocity wants to own. Formats include in-chat ads, in-chat display ads, multi-brand carousels, polls, loading-time ads, sponsored follow-up questions, and playable ads. That matters because the company is not simply selling a banner slot inside a chatbot. It is trying to standardize ad units for AI interfaces before app developers, ad networks, and model companies settle on their own defaults. For advertisers, Velocity is selling access to live intent. Its [advertiser page](https://velocity.io/advertisers) frames prompts as a new targeting signal alongside behavioral and contextual data. A user asking an AI assistant how to get back into running after a two-year break, for example, can be matched with a sponsored running gear placement inside the answer flow. That is the commercial opportunity and the risk in the same product surface. Search advertising worked because users often reveal intent at the moment they are ready to act. AI chats go further: users may disclose goals, health questions, personal constraints, anxieties, finances, work problems, or relationship issues in the same interface where Velocity wants to insert commercial recommendations. Velocity says its conversation manager can extract real-time intent and context signals, filter personally identifiable information, and route anonymized signals to demand sources. Its [privacy policy](https://velocity.io/privacy-policy), last updated June 17th, 2026, also says the company collects device and network data, advertising IDs, consent strings, inference data derived from end-user interactions with publisher properties, and advertising transaction data. The policy says publishers are responsible for getting user consent where required, while Velocity acts as a data controller for personal information processed through its services. That structure puts a heavy burden on implementation. If AI apps do not clearly disclose when a response contains a paid placement, or if users do not understand that their chat behavior may generate ad-targeting inferences, Velocity's infrastructure could inherit the same trust problem that has followed adtech across browsers and mobile apps. The company says its goal is to understand intent without exposing personal information. Its privacy policy shows the operational version is more complicated: Velocity, publishers, ad partners, exchanges, analytics companies, and demand-side platforms can all become part of the data chain. ### Early customers are real, performance claims are still company-supplied Velocity says early customers and partners include Leadtech / MAU, AIBY's ChatOn, Liner, Verve, DeepAI, Appier, 44pixels, VYBS, Kolbo.AI, Teads, Jetkite, AppCentral, Hiding Elephant, and Whtvr.AI. The company also says its technology is already used by AI companies serving millions of daily active users. Those adoption claims are meaningful, but they are still Velocity's claims. The company has not disclosed revenue, take rate, gross merchandise value, net revenue retention, ad load, conversion rates, or the share of prompts eligible for monetization. It has also not disclosed whether the $27 million seed round included a valuation. The customer quotes Velocity publishes focus on responsiveness, monetization results, user experience, and partner quality. They do not provide enough detail to judge whether native ads inside AI chats can become a durable revenue stream rather than an incremental yield tactic. That distinction is the round's central question. The ironSource comparison cuts both ways. ironSource helped define mobile app monetization and distribution before Unity acquired it in an all-stock deal in 2022. Unity later said in 2026 that it would sunset the ironSource Ads Network and pursue a divestiture of Supersonic, its game publishing business. That does not erase the founders' experience building mobile growth systems, but it does show how quickly ad infrastructure can become strategically expendable when platform economics shift. Velocity is trying to arrive before the AI app market hardens. The company currently has about 20 employees and operates an R&D center in Israel, according to CTech. Shoham told the publication Velocity had built its system with 15 people until shortly before the interview and could have raised more money, but chose to stay focused on execution. The timing is rational. AI app creation has become cheaper, distribution remains expensive, and inference costs punish products that attract heavy free usage without a path to paid conversion. A monetization layer that lets AI apps keep free tiers open could become important infrastructure if users tolerate ads inside conversational interfaces. The tolerance question will decide Velocity's market. Mobile ads were accepted because users understood the trade: free apps in exchange for interruptions, rewarded placements, or tracking. AI chats feel closer to private assistance. Velocity's founders know how to monetize scaled software usage. Their harder job is making the ad layer visible enough to be trusted and native enough that AI apps keep it turned on.