What it does
Before you submit, click Review with AI on the public form. Aclamos sends your draft answers to Claude, which compares them against the same rubric the judges will use. You get back:
- Per-criterion read — for each criterion in the rubric, is your draft
strong,weak, orunclear(didn't address it)? - Top 3 fixes — concrete, imperative actions like "Add a metric for impact in paragraph 2."
- One-paragraph summary — what you got right + your single biggest gap.
You always own the final submission. The AI doesn't auto-rewrite anything; it just tells you what a judge would notice.
Why it's not a ghost-writer
Two reasons:
- Honesty. Awards programs that prize originality penalize AI-generated copy. We coach, we don't write.
- Calibration. A judge reading your real voice + an AI-checked structure ranks higher than a judge reading polished AI prose that's missing the substance.
When it helps most
- First-time entrants. You don't know what judges look for; the rubric is a black box. Self-check makes the rubric concrete in your context.
- Half-finished drafts. You wrote 70% of the form, then froze. Run self-check; the "unclear" reads tell you exactly which paragraph to write next.
- Last-minute polish. 30 minutes before deadline. You don't have time to ask a friend to read it. Self-check is a 15-second second pair of eyes.
What it costs you
Free for entrants. The producer's AI credit pool pays — same metering as any other AI feature on Aclamos. Producers can disable AI features per show (entrants will see the button greyed out).
Privacy
Your draft is sent to Anthropic's Claude API for the duration of the call only. We don't store the draft text on Aclamos servers; we store an audit trail of "AI self-check ran on category X at IP Y for N characters." Same retention rules as the rest of the platform.
Rate limits
10 self-checks per hour per IP per show. If you hit the cap, finish the draft and submit — judges will read the same thing the AI would have.