AI use case open data
Download the whole library: use cases, value benchmarks and every public deployment record with its quote and source. Free to reuse under CC BY 4.0.
JSON
Full library
Use cases with every section, computed benchmarks, indicative values and all public evidence records.
CSV
Use cases
One row per use case with taxonomy, risk tier, evidence count and indicative value.
CSV
Deployments and metrics
One row per reported metric, with organization, value, claimant, exact quote and source.
License and citation
Licensed under Creative Commons Attribution 4.0. Cite as: Blits.ai AI Use Case Library, https://www.blits.ai/ai-use-cases, retrieved 2026-09-28. The data covers public records only; anonymized Blits.ai deployments are not included. Current size: 201 use cases, 666 deployment records, 521 organizations.
Data dictionary
| Field | Meaning |
|---|---|
| slug | Stable identifier of the use case; the page is https://www.blits.ai/ai-use-cases/{slug}. |
| industries, functions, patterns | Taxonomy ids: the industries, business functions and AI patterns the use case belongs to. |
| audience, autonomy | Who interacts with the AI, and how much it does without a human (assist, copilot, supervised agent, autonomous). |
| eu_ai_act_tier | Risk tier under the EU AI Act with its basis: minimal, limited (transparency), high, or depends on design. |
| indicative_value_low, _high | Worked example value for the reference organization, from the stated inputs and formula. |
| evidence_id | Stable identifier of a deployment record; also the anchor on the use case page. |
| stage | Deployment stage: announced, pilot, production, scaled, or paused. |
| grade | Evidence grade A to D, by type of source (see methodology). |
| kpi, value, unit, qualifier | The reported metric, its number and unit, and whether it was stated as exact, approximately, up to or at least. |
| claimant | Who made the claim: the organization, a vendor, a regulator or an independent party. |
| quote, source_url | The exact sentence stating the number, and where it was published. |
| verification, checked_at | How the record was checked, and when. |
Controlled vocabularies (industries, functions, AI patterns, KPIs, grades) are listed with definitions in the JSON file and on the methodology page.