Kevel Raises $52M to Scale AI Ad-Serving APIs

Kevel, the API-first ad-serving infrastructure company, has raised $52 million to scale its platform and expand AI-driven tooling aimed at the fast-growing commerce media segment. The round reinforces a broader market thesis: retailers and marketplaces increasingly want to own their advertising stack rather than rent it from a black-box SSP or a closed retail media network vendor.

Why an infrastructure play matters

Unlike full-stack ad platforms that bundle UI, reporting, and demand, Kevel sells the plumbing. Its APIs let a publisher, retailer, or marketplace build a custom ad server — decisioning, inventory management, forecasting, and serving — without constructing those systems from scratch. For engineering teams, that is the difference between shipping a retail media network in months versus years.

This "headless ad server" model mirrors the shift toward composable infrastructure seen elsewhere in ad tech. Where legacy ad serving centered on monolithic platforms like Google Ad Manager, the API-first approach gives product and ad-ops teams direct control over auction logic, targeting rules, creative rendering, and data flows. It is a meaningful alternative for companies that view their first-party commerce data as a strategic asset they do not want to expose to a third party.

The commerce media tailwind

Commerce media — the broader category encompassing retail media networks and marketplace advertising — remains one of the few high-growth zones in digital advertising. Retailers from grocers to travel platforms are standing up sponsored-product placements, on-site display, and increasingly off-site and connected TV extensions powered by their purchase data. The economics are compelling: ad revenue carries far higher margins than core retail, and first-party transaction data is durable in a post-third-party-cookie world.

That durability is central to Kevel's positioning. As signal loss from cookie deprecation and privacy regulation continues to erode open-web addressability, deterministic first-party commerce data becomes disproportionately valuable. An infrastructure provider that lets retailers activate that data on their own terms — without routing it through external identity graphs — sits in an advantageous spot.

The AI angle

The funding is explicitly earmarked for AI-driven capabilities. In an ad-serving context, that typically translates to automated campaign setup, yield optimization, predictive forecasting, creative assembly, and natural-language interfaces for ad-ops workflows. For a platform selling to retailers whose core competency is commerce rather than ad tech, lowering the operational burden through AI is a direct path to adoption — fewer specialized ad-ops hires required to run a sophisticated ad business.

The practical question for buyers is whether AI tooling improves the metrics that matter: fill rate, yield per impression, campaign pacing accuracy, and time-to-launch for new advertisers. Infrastructure vendors that can demonstrate measurable lift on those fronts will differentiate from the growing field of commerce media point solutions.

Implications for the video and programmatic ecosystem

While commerce media began on-site with sponsored products, its trajectory increasingly intersects with the open programmatic and CTV worlds. Retailers are extending their audiences off-site and onto connected TV, where their purchase data informs targeting and closed-loop measurement. Infrastructure that can serve across formats — including video and CTV ad pods via standards like VAST and server-side ad insertion — positions a vendor to capture that expansion.

For the broader programmatic landscape, Kevel's raise is another data point confirming that the ad-serving layer itself is being unbundled and recomposed. As more large advertisers and retailers build proprietary decisioning engines, the role of traditional intermediaries shifts. SSPs, DSPs, and exchanges will increasingly interoperate with a long tail of custom-built retailer ad servers rather than a handful of dominant platforms — raising the stakes for clean integration via OpenRTB, supply-chain transparency, and standardized measurement.

At $52 million, this is a sizable bet on the infrastructure thesis. Whether API-first ad serving becomes the default architecture for commerce media or remains a specialist choice will depend on how far the AI tooling lowers the barrier to entry — and how aggressively incumbent full-stack vendors respond.

Stay on top of video ad serving and programmatic. Follow Adelerate.