AI Prompt Generator Platform
Budget: ₹12,500 – ₹37,500 INR
The goal is to launch a web-based platform that automatically crafts and refines prompts for multiple large language models. A user should be able to provide a brief objective, choose the target model—OpenAI GPT-3, Google Bard, Anthropic Claude, or Gemini 3—and receive a well-structured prompt ready for use. I want the workflow to encourage intermediate-level input: the user supplies some context and a few key parameters, then the system completes the rest, returning both the prompt and a short rationale so the user understands why each element was suggested.
Core requirements
• Create new prompts from scratch based on user-supplied goals.
• Refine existing prompts, improving clarity, tone, and specificity without altering intent.
• Model-aware logic so the prompt formatting adjusts automatically for each LLM’s quirks.
• Clean, intuitive UI with fields for goal, persona, style, length constraints, and model selection.
• Modular code so additional models can be plugged in later.
• Export options: copy to clipboard, download as text/JSON, or save to user workspace.
Deliverables
1. Running web application (front-end + back-end) with the above functionality.
2. Source code and concise setup instructions.
3. A short usage guide illustrating how to create and refine prompts for GPT-3, Bard, Claude, and Gemini 3.
4. One revision round after initial hand-off to iron out usability issues.
Acceptance criteria: the system must generate a usable prompt for each supported model in under five seconds and improve an uploaded prompt’s evaluation score (simple rubric provided on delivery) by at least one grade.
Core requirements
• Create new prompts from scratch based on user-supplied goals.
• Refine existing prompts, improving clarity, tone, and specificity without altering intent.
• Model-aware logic so the prompt formatting adjusts automatically for each LLM’s quirks.
• Clean, intuitive UI with fields for goal, persona, style, length constraints, and model selection.
• Modular code so additional models can be plugged in later.
• Export options: copy to clipboard, download as text/JSON, or save to user workspace.
Deliverables
1. Running web application (front-end + back-end) with the above functionality.
2. Source code and concise setup instructions.
3. A short usage guide illustrating how to create and refine prompts for GPT-3, Bard, Claude, and Gemini 3.
4. One revision round after initial hand-off to iron out usability issues.
Acceptance criteria: the system must generate a usable prompt for each supported model in under five seconds and improve an uploaded prompt’s evaluation score (simple rubric provided on delivery) by at least one grade.