Validating a High-Performance Cloud Object Store with Lightweight Formal Methods
Summary
A paper on using lightweight formal methods (reference models, property-based testing, stateless model checking) to validate the distributed metadata index of Amazon S3 Express One Zone, showing the approach scales sustainably to production engineering.
Similar Articles
Making Failure Safe: A Constrained, Verifiable Agent Framework for Open-Web Data Collection
This paper proposes a constrained, verifiable agent framework for open-web data collection that shifts LLM output from free-form code to typed JSON collector configurations, achieving zero execution-stage LLM tokens and low latency on 80 tasks.
DocOps: A Verifiable Benchmark for Autonomous Agents in Complex Document Operations
This paper introduces DocOps, a deterministically verifiable benchmark for evaluating autonomous agents on complex document operations, revealing key failure modes such as long-term state tracking collapse, shallow semantic verification, and destructive editing of structural metadata.
Learning to Formalize: Bridging Open-Ended Workloads and Fast Decision Models
This paper proposes a small candidate formalizer trained on offline teacher-generated lists to bridge the interface gap between open-ended LLM workloads and fast decision models, reporting up to 52.6% task success on CALVIN-D with 2.09s end-to-end latency, compared to 11.8% without the formalizer and slower baselines like Qwen3.5-27B.
Formal Methods Meet LLMs: Auditing, Monitoring, and Intervention for Compliance of Advanced AI Systems
This paper proposes techniques that combine formal methods (Linear Temporal Logic) with LLMs for auditing, monitoring, and intervening in AI systems to ensure compliance with behavioral constraints, showing that even small-model labelers can match frontier LLM judges in detecting violations.
Cloud-Native Evaluation-as-a-Service: A Microservices Architecture for Scalable AI Monitoring with Conformal Guarantees
This paper presents EaaS, a cloud-native microservices architecture for scalable AI monitoring that provides conformal prediction, calibration assessment, drift detection, and fairness monitoring with statistical guarantees.