Programming with Data: Test-Driven Data Engineering for Self-Improving LLMs from Raw Corpora

TL;DR

This paper introduces a structured, feedback-driven approach to data engineering for large language models, enabling targeted repairs and systematic improvements by mapping data lifecycle to software development processes.

Contribution

It formalizes the Programming with Data principle, linking data management to software engineering, and demonstrates its effectiveness across multiple scientific disciplines with open resources.

Findings

Model failures can be traced to specific data deficiencies.

Targeted data patches improve model performance across scales.

Structured knowledge enables systematic data-driven model repair.

Abstract

Reliably transferring specialized human knowledge from text into large language models remains a fundamental challenge in artificial intelligence. Fine-tuning on domain corpora has enabled substantial capability gains, but the process operates without feedback: when a model fails on a domain task, there is no method to diagnose what is deficient in the training data, and the only recourse is to add more data indiscriminately. Here we show that when a structured knowledge representation extracted from the source corpus serves as the shared foundation for both training data and evaluation, the complete data-engineering lifecycle maps onto the software development lifecycle in a precise and operative way: training data becomes source code specifying what the model should learn, model training becomes compilation, benchmarking becomes unit testing, and failure-driven data repair becomes debugging. Under this correspondence, model failures decompose into concept-level gaps and reasoning-chain breaks that can be traced back to specific deficiencies in the data and repaired through targeted patches, with each repair cycle producing consistent improvements across model scales and architectures without degrading general capabilities. We formalize this principle as Programming with Data and instantiate it across sixteen disciplines spanning the natural sciences, engineering, biomedicine, and the social sciences, releasing a structured knowledge base, benchmark suite, and training corpus as open resources. By demonstrating that the relationship between training data and model behaviour is structurally traceable and systematically repairable, this work establishes a principled foundation for the reliable engineering of human expertise into language models.