AgentLTV: An Agent-Based Unified Search-and-Evolution Framework for Automated Lifetime Value Prediction

2026-02-25

AI Summary

This work addresses the challenge of lifecycle value (LTV) prediction, which typically relies on costly and non-transferable custom modeling pipelines tailored to specific decision scenarios. To overcome this limitation, the authors propose a unified automated machine learning (AutoML) framework powered by large language model (LLM) agents. The approach encodes candidate solutions as executable pipeline programs and employs a two-stage optimization strategy combining Monte Carlo tree search with island-based evolutionary algorithms. It further incorporates Pareto-aware multi-objective evaluation and bucket-level calibration validation. As the first study to integrate LLM agents with two-stage AutoML for LTV prediction, the method demonstrates significant improvements in ranking performance and calibration accuracy on both private large-scale datasets and public benchmarks, with particularly strong results for high-value and negative-LTV user segments, and has been successfully deployed in production.

Technology Category

• Machine Learning: Auto ML and Hyperparameter Tuning
• Natural Language Processing: (Large) Language Models
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Lifetime Value (LTV) prediction is critical in advertising, recommender systems, and e-commerce. In practice, LTV data patterns vary across decision scenarios. As a result, practitioners often build complex, scenario-specific pipelines and iterate over feature processing, objective design, and tuning. This process is expensive and hard to transfer. We propose AgentLTV, an agent-based unified search-and-evolution framework for automated LTV modeling. AgentLTV treats each candidate solution as an {executable pipeline program}. LLM-driven agents generate code, run and repair pipelines, and analyze execution feedback. Two decision agents coordinate a two-stage search. The Monte Carlo Tree Search (MCTS) stage explores a broad space of modeling choices under a fixed budget, guided by the Polynomial Upper Confidence bounds for Trees criterion and a Pareto-aware multi-metric value function. The Evolutionary Algorithm (EA) stage refines the best MCTS program via island-based evolution with crossover, mutation, and migration. Experiments on a large-scale proprietary dataset and a public benchmark show that AgentLTV consistently discovers strong models across ranking and error metrics. Online bucket-level analysis further indicates improved ranking consistency and value calibration, especially for high-value and negative-LTV segments. We summarize practitioner-oriented takeaways: use MCTS for rapid adaptation to new data patterns, use EA for stable refinement, and validate deployment readiness with bucket-level ranking and calibration diagnostics. The proposed AgentLTV has been successfully deployed online.

Problem

• Lifetime Value Prediction
• Scenario-specific Modeling
• Pipeline Automation
• Model Transferability
• LTV Data Patterns

Innovation

• Agent-based framework
• Automated LTV prediction
• Monte Carlo Tree Search
• Evolutionary Algorithm
• Executable pipeline program