Data lineage connects a source to its transformations and use. For an AI reviewer, that may include document version, extraction, chunking, indexing, retrieval, generation, and human disposition.
Data-integrity language is useful only when connected to an observable control. Ask whether a reviewer can identify the source, understand transformations, retrieve the relevant version, and distinguish generated text from approved information.
Worked example
Worked example — fictional
A PDF contains a table with the requirement in one column and an exception in another. Extraction merges the columns in the wrong order. The model interprets the exception as a separate mandatory rule.
To investigate, the team needs the original document version, extracted representation, relevant retrieved content, configuration, and generated finding. Keeping only the final answer makes the failure difficult to explain.
Practical evidence
Use representative document formats, scanned content where applicable, tables, footnotes, revision markings, and attachments. Compare extraction with source meaning. Define how unsupported formats are detected and handled rather than accepting incomplete extraction silently.
Completed record wording: “The evaluation record identifies the source document and revision, extracted-text artifact, retrieval configuration, generated findings, and reviewer dispositions. Retention and access are governed by the approved information and record assessment.”
Exercise
The system retains the source URL but the page changes over time. Can the original answer always be reconstructed?
Show suggested answer
Not necessarily. Evaluate versioned sources or an appropriate retained representation and metadata, consistent with permissions and record obligations. A mutable address alone may not establish the content used.
Expected records
Source and transformation identifiers, extraction checks, access rules, data-quality findings and the approved retention approach. Adapt record names and detail to the applicable procedures; these are practitioner examples, not a universal mandatory document list.
Relevant regulations and guidance
Regulations set requirements within their scope; guidance offers recommendations or explains applicable expectations; standards depend on adoption or incorporation. Voluntary frameworks do not establish compliance. The lesson's controls, test designs and example records are practitioner suggestions unless explicitly identified otherwise.
Regulation · U.S. Food and Drug Administration · Current eCFR text
Scope: United States · Electronic records and signatures within FDA Part 11 scope, tied to the underlying required records.
Relevance to this lesson: For records in scope, assess integrity, access and retrievability rather than assuming a source URL is sufficient.
Publication status: Final / published. Source checked 21 September 2026.
Publication identity, scope and status checked against the linked official source. This is not an independent audit of every retained guide statement.
Source details and applicabilityGMP guideline · European Commission · January 2011 revision
Scope: European Union · Computerised systems supporting activities in the applicable EU pharmaceutical GMP context.
Relevance to this lesson: Provides computerised-system context where the relevant GMP guideline applies; detailed section interpretation requires authoritative text.
Publication status: Final / published. Source checked 21 September 2026.
The Commission index confirms the January 2011 operative revision. The linked PDF could not be retrieved in this check; retained detailed section mappings need review against an authorized copy.
Source details and applicability