{"stable_id":"4a77ab1ce686bbd8","slug":"google-llm-code-migration","company":"Google","workflow_name":"LLM-assisted 32-bit to 64-bit identifier migration","function_code":"software_engineering","pattern_codes":["creator_to_judge"],"changed_assumption":"Large migrations can combine machine-found locations, LLM edits, automated checks, and engineer review.","evidence_strength":"verified","publication_tier":"showcase","freshness":"current","reviewed_at":"2026-08-23","updated_at":"2026-08-22","source_quality_summary":"1 peer reviewed; publication outcomes are verified.","caveat_summary":"Experience report without a randomized comparator. Time reduction is the developers' estimate, not instrumented time measurement.","freshness_caveat":null,"workflow_summary":{"before":[{"actor":"Software engineer","action":"Finds migration sites and writes context-specific edits across the monorepo.","actor_type":"human"},{"actor":"Software engineer","action":"Repairs failures and submits each change list.","actor_type":"human"}],"hinge":"Large migrations can combine machine-found locations, LLM edits, automated checks, and engineer review.","after":[{"actor":"Migration tooling","action":"Discovers likely change locations and prompts a monorepo-trained LLM to generate edits.","actor_type":"system"},{"actor":"Software engineer","action":"Reviews, repairs where needed, and submits validated change lists.","actor_type":"human"}],"decision_mode":"shared","decision_marker":"Human moves from creator to judge"},"before":[{"order":1,"actor":"Software engineer","action":"Finds migration sites and writes context-specific edits across the monorepo.","handoff_to":"Build and test systems","control":"Manual code review and testing."},{"order":2,"actor":"Software engineer","action":"Repairs failures and submits each change list.","handoff_to":"Code reviewer","control":"Reviewer approval."}],"after":[{"order":1,"actor":"Migration tooling","action":"Discovers likely change locations and prompts a monorepo-trained LLM to generate edits.","handoff_to":"Automated validators","control":"Location discovery, compilation, tests, and static checks."},{"order":2,"actor":"Software engineer","action":"Reviews, repairs where needed, and submits validated change lists.","handoff_to":"Code reviewer","control":"Human review and standard submission controls."}],"decision_rights":"Automated checks may reject edits; engineers decide whether to repair and submit; reviewers retain merge approval.","exception_path":"Failed or semantically difficult edits return to engineers for manual correction or conventional migration tooling.","removed_work":["Most routine edit authoring","Repeated context reconstruction at known migration sites"],"outcomes":[{"metric":"LLM contribution to accepted migration work","baseline":"Traditionally manual, context-dependent migration edits.","result":"74.45% of 595 submitted code changes and 69.46% of 93,574 edits were LLM-generated.","period":"Twelve-month case study.","scale":"39 migrations completed by three developers.","attribution_caveat":"Experience report without a randomized comparator.","evidence_label":"verified"},{"metric":"Estimated migration effort","baseline":"Earlier manual migrations.","result":"Developers estimated a 50% reduction in total time.","period":"Twelve-month case study.","scale":"39 identifier migrations.","attribution_caveat":"Time reduction is the developers' estimate, not instrumented time measurement.","evidence_label":"verified"}],"executive_lesson":"The reusable pattern is not autonomous coding; it is AI generation inside a deterministic discovery, validation, and review envelope.","anti_pattern":"Allowing model-generated repository changes to bypass compilation, tests, or human review.","questions_for_leaders":["Which migrations are variable enough to benefit from LLMs?","What evidence must pass before a change reaches review?","How will estimated savings be instrumented?"],"collections":["human-still-decides"],"bundle_version":"1.0.0","bundle_fingerprint":"sha256:c23c6cc2b88153d008ea8fda928f632ce0011fc2d4c5036672a16e5d895bab93","canonical_url":"https://brianletort.ai/transformations/google-llm-code-migration"}