Write it like you would to a friend. We'll make it a rubric.

Cost per call:₹0 · In-house inference

Most hiring managers write briefs in paragraph form. Most rubric builders demand weighted dimensions with descriptor bands. We bridge the two — you write naturally, we propose the structure, you confirm or edit, we lock.

API endpoint

POST /api/rubrics/generate

Model routing

In-house specialized inference

Cost envelope

~₹0 per extraction

Before · the paragraph you'd actually write

"We need someone to audit our checkout funnel and come back
with ideas. They should know a11y reasonably well, be good with
Figma, and actually talk to 5 real users before proposing anything.
If the proposal doesn't lead to a measurable drop-off reduction,
it's a no. Tier-1 MBA not required - portfolio is what matters."

After · the structured rubric we propose

{
  "dimensions": [
    { "name": "research_rigor", "weight": 0.25, "rubric": {
        "90-100": "Talks to 5+ real users with recorded sessions, triangulates findings against analytics or funnel data, names 3+ falsifiable hypotheses",
        "70-89":  "Talks to 3-4 real users, surfaces 2+ hypotheses...",
        "40-69":  "Desk research only; generic hypotheses...",
        "0-39":   "No user research evidence"
    }},
    { "name": "visual_craft", "weight": 0.30, ... },
    { "name": "accessibility", "weight": 0.20, ... },
    { "name": "measurable_claim", "weight": 0.25, ... }
  ],
  "confidence": 0.86,
  "unresolved_questions": [
    "Target drop-off reduction threshold not specified — suggest 20%",
    "Accessibility bar (WCAG AA vs AAA)?"
  ]
}

Write a brief, get a rubric