The advertising and marketing industry has experienced rapid speed from digitalization, and artificial intelligence is now enabling autonomous operations at scale. For companies developing next-generation advertising and marketing technologies, the challenge is no longer whether to adopt AI but whether their infrastructure can handle the speed and scale the industry demands.

Alembic, a causal AI platform, addresses a critical business challenge: proving which marketing initiatives actually drive growth rather than simply reporting what happened. Using NVIDIA DGX Vera Rubin NVL72 systems, Alembic scales its models to analyze more variables, run larger simulations and quantify true drivers of growth across marketing investments. Alembic will be the first causal AI company to use NVIDIA DGX Vera Rubin SuperPODs for enterprise-scale modeling, providing executives a single source of unbiased truth about what drove business outcomes and where capital is being wasted. The infrastructure runs on private supercomputing inside Equinix data centers, keeping AI workloads local, with World Wide Technology extending this to secure and regulated environments.

Amazon Web Services combines cloud infrastructure, foundation models and NVIDIA GPU-accelerated computing for the adtech industry, enabling AI agents at scale. AWS provides advertisers, demand-side platforms, supply-side platforms and software vendors a production-ready reference implementation for AI-powered bidding directly within auctions using NVIDIA Triton Inference Server, moving companies from rules-based decisions to AI-powered models for bid optimization, audience activation and deal scoring within live auction pipelines.

Criteo, which shows the right product to the right shopper at the right moment across one of the largest recommendation networks in digital advertising, achieved roughly two times speedup in model training on NVIDIA Blackwell GPUs using the NVIDIA cuEmbed open library. This efficiency frees approximately seventeen thousand GPU hours yearly, with further scaling underway.

Taboola applies similar infrastructure to conversational AI through DeeperDive, its AI answer engine powered by NVIDIA GPUs, extending this to AI platforms and chatbots for revenue generation through advertising. Higgsfield AI, a production platform for AI video and image generation, uses NVIDIA infrastructure to power Higgsfield Supercomputer agents managing the complete marketing automation lifecycle from campaign ideation through autonomous optimization in a single interface. The platform orchestrates leading large language models alongside thirty-five-plus image, audio and video models including Higgsfield's proprietary Soul and Soul 2.0 models built on NVIDIA Blackwell. Marketing campaigns for nearly four hundred Fortune 500 companies are created on the platform.

KERV.ai, an AI-powered media leader, optimized its processing pipeline using the NVIDIA Nemotron 3 Nano Omni open model, achieving over ten times improvements in speed and efficiency. Its Moment Match Engine evaluates multiple signals across video frames and media assets to provide content recommendations based on ad creative. On MediaPerf, an open benchmark for AI video understanding, Nemotron 3 Nano Omni delivered the highest throughput and lowest inference cost of any model evaluated, open or closed source.