{"schema_version":1,"assessment_id":"eoa_inverse_innovation_exp03_opportunity320_20260801","source_experiment_id":"eoa_inverse_innovation_exp03_full320_20260801","cell_id":"layer_decay_and_expiration_management__data_science","archetype_slug":"layer_decay_and_expiration_management","domain_slug":"data_science","title":"Governed lifecycle management for stale analytical artifacts","opportunity_summary":"A lifecycle service would inventory versioned analytical artifacts, identify suspected supersession, apply retention and dependency gates, demote eligible artifacts from active discovery through reversible archive or quarantine, and verify restoration. The candidate is structurally testable and safety-bounded, but the prevalence of the stated problem, operational effect, lineage quality, restoration reliability, and distinctiveness remain unverified.","adopter_authorizer":"Data and ML platform operators are the likely operational adopters. Artifact owners may propose refresh or demotion, while data governance, legal, privacy, compliance, audit, and model-risk authorities authorize sensitive, contested, held, or irreversible dispositions.","scores":{"meaningful_impact":{"score":4,"rationale":"If superseded artifacts are being reused or retained because cleanup is unsafe, reducing their discovery prominence while preserving reconstruction could improve analytical currency, control accumulation, and reduce destructive cleanup risk. The sealed candidate does not establish how often or how severely this problem occurs."},"stakeholder_pull":{"score":3,"rationale":"The proposal identifies concrete burdens for analysts, platform operators, artifact owners, and control functions, including stale discovery, cleanup avoidance, and audit or rollback concerns. No observed demand, adoption request, or stakeholder commitment is supplied."},"incremental_advantage":{"score":4,"rationale":"Relative to owner-by-owner cleanup and a badge-only metadata catalog, the composition adds retention authority, dependency-gated disposition, reversible archive or quarantine, restore drills, tombstones, and rollback. Whether these additions improve outcomes enough to justify their burden is untested."},"distinctiveness_plausibility":{"score":3,"rationale":"The integration of discovery demotion, policy authority, dependency checks, reversible disposition, and tested restoration is coherently differentiated from the stated nearest rival. Prior art is explicitly unsearched, so real-world distinctiveness cannot be verified closed-book."},"technical_implementability":{"score":3,"rationale":"A shadow workflow using inventories, labels, simulated dispositions, archive samples, and restores appears technically plausible. Feasibility depends materially on identity resolution, lineage completeness, durable lifecycle metadata, archive integrity, and integration across heterogeneous artifact types."},"adoption_authority_feasibility":{"score":4,"rationale":"The candidate assigns operational and governance authority by reversibility and sensitivity, blocks unresolved dependencies and holds, and provides halt and rollback conditions. Cross-functional approvals and contested ownership could still slow adoption."},"evidence_readiness":{"score":4,"rationale":"A specified 30-day non-production shadow pilot can compare lifecycle handling with ordinary cleanup and can measure search visibility, triage effort, false retirements, exceptions, dependency issues, and restoration. No pilot data or validated measurement protocol is yet supplied."},"safety_net_benefit":{"score":4,"rationale":"Reversible tiering, quarantine, dependency blocks, holds, audit records, tombstones, restore drills, and explicit rollback directly reduce the risk of erasing required evidence or breaking active work. Incomplete lineage and improper retention remain serious residual hazards."},"scalability":{"score":3,"rationale":"Registry-based policies and periodic revalidation could extend across many artifact classes, but semantic staleness is not uniform and each class may require different retention rules, dependency models, owners, restore procedures, and compliance handling."}},"score_confidence":"MODERATE","costs":{"first_evidence":{"band_2026_usd":"50K_TO_250K","scope":"A 30-day shadow pilot for one non-production artifact class, including inventory and identity resolution, stale-label review, simulated policy decisions, a stratified archive-and-restore sample, baseline comparison, safety review, and results analysis.","confidence":"MODERATE","assumptions":["Existing registry and archive capabilities can be reused.","The pilot does not include hard deletion or production automation.","A limited number of platform, analyst, governance, privacy, and audit participants are needed.","Data access and compliance review remain bounded to one artifact class."]},"initial_deployment_startup":{"band_2026_usd":"250K_TO_1M","scope":"Design and implementation of an initial production-capable lifecycle service for a small number of artifact classes, including policy configuration, dependency checks, reversible disposition, discovery integration, audit logging, restoration procedures, controls, and training.","confidence":"LOW","assumptions":["Core catalog, lineage, identity, storage-tiering, and access-control infrastructure already exists.","Significant integration and metadata remediation are required but no platform replacement is undertaken.","Irreversible automated deletion remains outside the initial scope.","Governance authorities can agree on retention classes and escalation paths."]},"operational_launch":{"band_2026_usd":"1M_TO_5M","scope":"Broader organizational launch across multiple analytical artifact classes and teams, including connector development, lineage improvement, policy harmonization, migration, security and compliance validation, restore testing, owner onboarding, support, and controlled rollout evaluation.","confidence":"LOW","assumptions":["Artifact systems and retention obligations are heterogeneous.","Production reliability, incident response, and audit readiness are required.","Historical ownership and dependency gaps need remediation.","Launch remains within one organization rather than a multi-enterprise product rollout."]},"annual_recurring":{"band_2026_usd":"250K_TO_1M","scope":"Annual operation of the lifecycle service, including infrastructure, archive storage and retrieval, policy and exception review, metadata stewardship, restore drills, monitoring, incident handling, support, compliance evidence, and periodic outcome evaluation.","confidence":"LOW","assumptions":["The service covers several artifact classes but not every possible analytical system.","Human review remains necessary for contested, sensitive, or irreversible cases.","Exception aging and lifecycle metadata require continuing stewardship.","Storage savings, if any, are not credited against the resource band."]}},"research_burden":"HIGH","earliest_credible_horizon":"0_TO_3_MONTHS","pipeline_gates":{"recognizable_externally_supportable_problem":{"status":"YES","reason":"The sealed candidate specifies externally checkable registry conditions: superseded artifacts remaining active, uncertain ownership and dependencies, inconsistent retention, stale discovery results, and absent demonstrated restoration. Its prevalence and severity still require measurement."},"identifiable_adopter_or_authorizer":{"status":"YES","reason":"Data and ML platform operators can adopt the service; artifact owners can propose reversible changes; and governance, legal, privacy, compliance, audit, or model-risk authorities can approve sensitive, contested, held, or irreversible actions."},"distinct_testable_incremental_claim":{"status":"YES","reason":"Against owner-by-owner cleanup and badge-only cataloging, the candidate claims that policy-authorized, dependency-gated, reversible disposition with restore verification will reduce stale artifacts in ordinary discovery or owner triage burden without unacceptable false retirements, broken dependencies, restore failures, or compliance exceptions."},"bounded_next_evidence_step":{"status":"YES","reason":"The candidate authorizes a 30-day shadow pilot on one non-production artifact class with simulated policy decisions and archive-and-restore testing of a stratified sample, while prohibiting hard deletion."},"no_unresolved_safety_or_authority_stop":{"status":"YES","reason":"The initial step is reversible and excludes automatic hard deletion. Unresolved dependencies, policy conflicts, legal holds, sensitive dispositions, failed restores, missing audit records, and unauthorized exposure trigger blocking, halt, or rollback rather than requiring unresolved authority."},"implementation_cost_scope_and_range":{"status":"YES","reason":"The candidate defines a bounded pilot and enough functional components to frame broad resource-equivalent ranges, although actual costs remain uncertain because system heterogeneity, metadata quality, and integration scope are unknown."}},"blocking_evidence":["A representative registry audit establishing whether superseded-but-active artifacts, stale-artifact reuse, and cleanup avoidance are materially present.","Baseline measures for stale artifacts appearing in ordinary discovery, owner triage time, retention-rule consistency, ownership coverage, and inbound-dependency coverage.","Pilot evidence on stale-label precision and harmful false retirement or demotion rates.","Demonstrated archive and restore success for a stratified sample, including retrieval time and reconstruction fidelity.","Counts and causes of policy conflicts, legal or privacy exceptions, unresolved dependencies, and exception aging.","Evidence that lifecycle handling improves search visibility or triage outcomes relative to ordinary cleanup without unacceptable safety or compliance failures.","Prior-art research comparing the composition with model registries, metadata catalogs, records-retention systems, data lifecycle management, and reproducibility platforms."],"next_evidence_step":"Run the authorized 30-day shadow pilot on one non-production artifact class. Compare lifecycle labels and simulated dispositions with the existing owner-by-owner baseline on stale artifacts appearing in ordinary search, owner triage time, dependency and ownership coverage, exception rates, and audit completeness; archive and restore a stratified sample. Falsify progression if benefits are absent or if false retirements, dependency breaks, restore failures, unauthorized exposure, or compliance exceptions exceed predeclared acceptable thresholds.","research_questions":["How many representative artifacts are superseded but still active, and how often do they appear in ordinary discovery or get reused as current evidence?","What share of candidate artifacts has an identifiable owner, complete inbound dependencies, an applicable retention class, and no unresolved hold?","How accurately can suspected staleness and supersession be identified without treating age as evidence of worthlessness?","Does reversible demotion reduce stale search exposure or owner triage time relative to ordinary cleanup?","What restoration success, reconstruction fidelity, and retrieval latency are achieved across a stratified archive sample?","How often do governance exceptions occur, why do they occur, and do they expire or accumulate as permanent holds?","What ongoing metadata stewardship and cross-functional review workload does the mechanism create?","Is the proposed composition meaningfully distinct from existing lifecycle, catalog, registry, retention, and reproducibility approaches?","Which artifact classes can share lifecycle policies, and which require class-specific semantics and controls?"],"recommendation":"PILOT_NOW","uncertainty_constraints":["Problem prevalence, severity, and stakeholder demand are not established by the sealed candidate.","No realized effect on discovery quality, triage effort, storage burden, reproducibility, compliance, or rollback has been observed.","World novelty and competitive distinctiveness are unmeasured.","Lineage completeness, ownership coverage, lifecycle-metadata durability, and archive reliability are unknown.","The proposal cannot infer semantic invalidity or deletion safety from age alone.","Cost bands are broad resource-equivalent estimates based only on the described scope, not verified implementation data.","Scaling across heterogeneous datasets, features, models, notebooks, and reports may require materially different policies and integrations."],"closed_book_prior_art_boundary":"Prior art is unsearched and unverified. This closed-book assessment makes no claim about novelty, market prevalence, existing implementations, or differentiation beyond the candidate's internally stated baseline and nearest rival; external prior-art research is required before any distinctiveness claim."}