Sequen

Sequen develops a behavior design engine and ranking platform that provides dynamic in-session personalization for consumer applications in the post-cookie, privacy-first AI era. The platform uses Large Event Models (LEMs), invented by Sequen, to generalize consumer behavior on sparse data, enabling personalization without requiring billions of users. Sequen is the first company to implement continual learning on dynamic user embeddings at inference, allowing real-time adaptation to user intent and preferences during active sessions. The platform operates with ultra-low latency of sub-20 milliseconds at p99, powering seamless in-session user experiences. Sequen's multi-horizon optimization handles consumer journeys that span multiple sessions and time windows, optimizing for both in-session and cross-session outcomes. The platform serves four primary use cases: product discovery, where micro-interactions dynamically optimize application surfaces for conversion; recommendations that interpret in-session intent and behaviors to deliver tailored suggestions beyond simple similarity matching; search results that dynamically re-rank as user behavior indicates preferences; and advertising that leverages consumer intelligence to improve paid goals alongside organic outcomes. Sequen enables consumer applications to become dynamic surfaces personalized for each user, moving beyond static profile attributes to leverage real-time behavioral signals. Customers report approximately 10% impact from in-session personalization, meaningfully shifting their mental model of personalization value. Founded in 2024 and headquartered in Long Island City, New York, Sequen has raised $22.0M in Series A funding from investors including White Star Capital, Threshold, Greycroft, VINYL, and Correlation Ventures.

AI Infrastructure > Enabling Technologies > Software > Data Preparation

College Wise > New York University