Create intermediate Works with: Any AI

Model Documentation Card

Checklist Has example

Document what a model is for and where it fails.

The Open buttons send the prompt text as currently shown below to that site only when you click. Fill in the placeholders first.

[VERIFY] is an editorial flag, not a blank to fill. Wherever the prompt shows [VERIFY], check that item against a reliable source before you use it.

Write a model card. Model name and purpose: [MODEL_AND_PURPOSE]. Training data and time period: [TRAINING_DATA]. Evaluation results overall and by group: [EVALUATION_RESULTS]. Intended users and decisions: [INTENDED_USE].

Use these sections and keep each short: summary (3 sentences); intended use and users; out-of-scope uses; data (source, period, size, labeling, known gaps); features and exclusions (and why); training procedure and version; evaluation (metrics with intervals, test set description, subgroup results in a table); limitations and failure cases with examples; ethical and legal considerations (people affected, protected characteristics, consent) [VERIFY in [JURISDICTION]]; monitoring and retraining plan; contact and owner; change log.
Add a risk table: Risk | Likelihood | Impact | Mitigation | Owner. Include a one-line plain-English warning at the top for users, and a checklist for review before deployment. Flag any information you do not have as [VERIFY] rather than inventing it.

Keep each section to a few sentences, link to the evaluation notebook, and date the card so readers know which model version it describes.

Fill in before use

Optional: type your details and the prompt above updates. Nothing is saved or sent.

Replace each bracketed placeholder with your own details, then paste into Any AI or another assistant.

Model: ticket priority classifier v1.2.
Purpose: suggest priority for new support tickets; humans decide.
Training data: 40,000 tickets, Jan 2024-Jun 2025, English only.
Evaluation: macro F1 0.74; high-priority recall 0.81; lower for tickets under 10 words (0.62).
Limitations: not tested on other languages; may misclassify sarcasm.
Intended use: triage assistance. Out of scope: automatic closing of tickets.
Monitoring: weekly agreement rate with agents; retrain when it falls below 70%.
Owner: Support analytics. Contact: analytics@example.org.
Update history: v1.1 added a spelling normaliser; v1.2 retrained with three more months of tickets. Review date: every quarter.

• model-card