Document AI systems retain memory across sessions for better accuracy
Amnesia by Design, Memory By Necessity: Persistent State for Document Intelligence
Computer Vision and Pattern RecognitionArtificial IntelligenceComputation and LanguageInformation RetrievalMachine Learning
Summary
Document AI systems today read and analyze documents but forget everything once done, starting fresh each time. This forgetting is a design choice, not a technical limitation, creating a barrier called the statelessness bottleneck. The authors explore how keeping track of document information over time can help systems improve understanding and catch contradictions. They propose a framework for managing this permanent memory and highlight how current tests miss evaluating it. Their work points to a future where Document AI remembers past documents and sessions to work smarter.
What this means in practice
- •For enterprise software developers: Build contract review tools that retain previous document changes to detect contradictions automatically over time.$Commercial implications: Enables development of smarter contract management products that maintain knowledge across document versions for enterprises.
- •For legal technology teams: Integrate persistent state frameworks to improve amendment tracking and reasoning in legal document analysis.
A survey. It maps existing work.
Authors
Souhail Bakkali, Ayoub Merimi
Abstract
Modern Document AI reads contracts, extracts fields, reasons over tables, and grounds answers to page regions, then forgets everything. Processing an amendment the next day begins from scratch: no schema retained, no contradiction detected, no experience carried forward. This is a structural choice, not a scale failure: current systems are stateless functions. We call this the statelessness bottleneck. This bottleneck lies beyond parameter scaling, context extension, and retrieval augmentation: storage provides persistence and retrieval provides access, but neither consolidates observations into knowledge that improves future processing. This survey formalizes persistent evidence-grounded document state as a unifying framework, specifying the operations and invariants required to convert multimodal evidence into durable, provenance-linked state. We introduce a statefulness audit showing that ten representative benchmarks, coded against eight statefulness criteria, leave cross-session state evolution untested, and derive a longitudinal benchmark harness with five counterfactual metrics: Experience Gain, Cost Efficiency, Memory Harm, Forgetting Fidelity, Coverage Retention, to characterize the benefit, cost, risk, and governability of persistent document state. Document AI lacks mechanisms coupling persistent state to document-native structure, provenance, and temporal validity. The next era of Document AI will be defined by what systems retain across documents, sessions, and time.