This paper outlines a novel architecture blueprint that use large language models (LLMs) to enhance ad targeting effectiveness through personalized messaging.
In 2017, Netflix demonstrated the impact of personalized promotional messaging by adapting artwork based on user preferences and viewing history. The company’s content experts generated multiple images for every title and change them regularly to lure audiences based on their previous viewing history. The approach utilized a sophisticated “online reinforcement learning” strategy, optimizing the balance between exploiting known user preferences and exploring new data for improved recommendations. This dynamic methodology was essential in minimizing the cumulative “regret” (defined as the difference between the expected “payoff” (e.g. engagement) of the algorithm and the payoff of a single fixed strategy for selecting artworks) and enhancing viewer satisfaction over time.
Despite its efficiency, this approach epitomized a few key principles under the legacy Personalization 1.0 paradigm, namely the need to choose between a finite set of creatives (while being limited by multimodal content creation’s cost and complexity), the reliance on high-quality (and therefore expensive) human preference labels to optimize algorithmic tuning, and the use of shallow context information (e.g. transient signals like search history) as proxies for viewers’ interests. Today’s advancements in Generative AI necessitate a complete re evaluation of the algorithmic / architectural trade-off and its relevance.
This paper outlines...
Exclusive Content
This article is available with a Technical Paper Pass
.jpg)
Selective multi-pass encoding for cost-effective video streaming
Tech Papers 2026: This paper presents a content-adaptive strategy, CASE, that predicts whether additional encoding passes would provide meaningful gains using a lightweight mechanism that derives spatial and temporal features from each video segment.
.jpg)
Scalable SSIM estimation from PSNR for per-title and context adaptive encoding workflows
Tech Papers 2026: This paper proposes ApproxSSIMate, a low-complexity method for estimating SSIM from PSNR combined with reference-sequence statistics.
.jpg)
Deep learning super resolution for dense dynamic point Cloud compression
Tech Papers 2026: This paper proposes Video-based Super Sampling Point Cloud Compression (VSS-PCC), a method that uses neural super-resolution to reduce the size of point cloud data before compression.
.jpg)
Client device discovery: An open standard for bridging the first-mile gap in cloud media workflows
Tech Papers 2026: This paper presents the architecture, protocol design, and security model of TR-12, demonstrates its contract-first approach using Smithy models, and discusses early implementation experience with the open source SDK.
.jpg)
Scene-adaptive camera: Dynamic imaging parameter control for localised shooting areas
Tech Papers 2026: This paper examines a scene-adaptive camera that dynamically adjusts imaging parameters across different areas of a frame to improve image quality.
