API engineering
Structured fact data: a provenance-first JSON design
A practical data model for claims, sources, evidence relationships, assessment history, and clearly defined uncertainty.
04 FIELDNOTES / TOPIC
Fact APIs
A focused reading path through The Evidence Journal. Interfaces for claims, records, and evidence should say exactly what they return. Explore the difference between finding information and assessing it, then connect that distinction to extraction, schemas, and service evaluation.
How to use this collection
These articles are most useful when designing a response contract or comparing an integration. Keep operational failures separate from factual assessments, preserve the original question, and require a source relationship that another reviewer can reconstruct. The API name alone should never imply that every returned statement has already been verified.
Start with the Fact API guide, then follow the fieldnotes below. All examples are illustrative and each article points to its own editorial source.
LLM evaluation
Choosing an AI LLM fact-checking service: a practical evaluation
Compare evidence boundaries, evaluation results, privacy questions, review effort, and integration requirements.
API engineering
Fact extraction APIs: turn text into atomic claims
Split complex text into checkable statements without losing the qualifiers, attribution, and context that change the question.
API engineering
Fact API fundamentals: from claims to verifiable evidence
Understand what a fact API should return, where evidence fits, and how to design a response that can be inspected.
Follow the claim. Find the evidence.
Go from a better question to a clearer evidence trail—one fieldnote at a time.