Intermediate Guide: Fix No Consistent Feedback Framework for Startup Curriculum Developers Using ChatGPT

You are a senior curriculum design specialist with 11 years of experience building assessment frameworks and feedback systems for fast-growing startup education companies. Help me build a consistent feedback framework so I can save 5 hours per week on assessment preparation and stop producing exam questions that vary wildly in quality across courses. My situation: - Type of courses I develop: [ONBOARDING TRAINING / TECHNICAL SKILLS / SOFT SKILLS / COMPLIANCE / MIX] - Current exam question process: [WRITTEN FROM SCRATCH EACH TIME / ADAPTED FROM TEMPLATES / NO FORMAL PROCESS] - Number of courses in development at once: [NUMBER] - Team size contributing to assessments: [JUST ME / 2-3 PEOPLE / LARGER TEAM] - Biggest feedback inconsistency: [e.g., "different reviewers give contradictory scores" / "no rubric for written responses" / "feedback too vague to act on"] - LMS or assessment tool used: [Typeform / Google Forms / Teachable / Custom / Other] Deliver: 1. A feedback framework structure — a 4-tier system covering immediate response feedback, formative assessment feedback, summative feedback, and learner self-assessment, with a template for each 2. An exam question quality rubric — a 6-criterion scoring guide that any team member can apply to evaluate whether a question is ready to publish 3. A question bank filing system — a folder and tagging structure for storing approved questions so they can be reused across courses without rewriting 4. A feedback language guide — 20 sentence starters organized by feedback type (encouragement, correction, extension) that produce consistent tone across reviewers 5. A 5-minute feedback review process — a checklist a curriculum developer runs before publishing any assessment to catch the most common quality issues 6. A time audit template — a breakdown of where assessment preparation hours currently go and the specific steps the framework eliminates 7. A feedback calibration exercise — a 30-minute activity for aligning two or more reviewers on scoring standards before a course launches Prioritize outputs that can be implemented in the current course sprint without requiring a full workflow rebuild.