Sponsors (in alphabetical order)

Agentic Infrastructure for Adaptive Customer Experience

Personalisation that spans every touchpoint

Speakers: Sami Abboud, Olivier Jeunen, Ellie Hanna
Attendees: Paul Meinshausen, Terry Lucy, Vineesha Raheja, Amaan Kulatunga

Consumers live in an attention economy—every brand is competing to “rent” a moment of a user’s time and deliver value in the most effective, efficient, and individualized way possible. Traditionally, engagement has depended on human teams manually segmenting users, defining rules, and orchestrating journeys that aim to feel personalised. But these teams are constrained by tooling—there’s a complexity ceiling that no amount of manual effort can break through. Adding to that complexity is the challenge that Data Science teams face, which is the gap between building models and seeing them used effectively in production. They spend disproportionate time on repetitive reporting, integration, and deployment tasks and not on higher-value research and experimentation. At the same time, access to clean, user-level behavioral data for feature engineering remains a persistent bottleneck.

Agentic infrastructure breaks down these challenges. It equips Marketing, Product, and Data teams with reinforcement learning agents that make real-time engagement decisions at scale. These agents continuously run thousands of parallelized experiments, learn from behavioural data, and dynamically adapt engagement variables like content, timing, and channel based on each individual’s preferences.

Many teams think of Aampe as an Agentic Experimentation System: a foundation that brings Marketing, Product, and Data teams together by enabling coordinated experimentation and tracking at the user level. CRM teams use Aampe to personalise outbound communications and uncover user preferences. Product teams use it as a user research engine, to experiment with in-app experiences.

Data teams can generate new data by feeding existing models (e.g., churn or voucher models) as priors into the agents and fine-tune their behaviour by adjusting reward weights. Agentic Infrastructure not only unifies these teams— it creates a living experimentation layer across the entire customer experience.

With Aampe, Data Scientists specifically benefit in the following ways:

  • Less time wasted on repetitive reporting and production churn through programmatic multi-model competition
  • Faster realization of impact from models through automated model deployment and rollback
  • Ability to focus on novel methods and research by deploying governance and audit trails
  • Ability to run more ambitious experiments without needing additional staff by using causal and counterfactual evaluation data
  • Greater leverage over the full ML lifecycle, not just model training through routine reporting and analytics automation

To learn more, visit us at www.aampe.com.

Diamond Supporter
 
 
 
Platinum Supporter
 
Gold Supporter
 
Bronze Supporter
 
 
 
Challenge Supporter
 
Women in RecSys’s Event Supporter
 
Breakfast Symposium
 
Coffee Break Sponsor
 
Special Supporters
 
 
 

This event is supported by the Capital City of Prague