AI Text Humanizer Prototype
Budget: $30 – $250 AUD
I need a lightweight AI-powered application that takes raw text—emails, social captions, technical write-ups, anything—and rewrites it so it sounds naturally written by a human.
Core objectives
• Primary function: improve text to sound more human.
• When the input is a technical document, the output must read in a clearly simplified tone that cant be detected by ai.
• For all other text types, the tool should keep the original intent while adding natural flow, contractions, and everyday vocabulary.
I would also like a % ai use
Scope for this phase
I’m looking for a working prototype that:
1. Accepts plain-text input (web form or simple desktop interface—your choice).
2. Uses an LLM (OpenAI, Hugging Face, or similar) with prompt engineering or fine-tuning to perform the humanization.
3. Returns the rewritten text instantly and displays it side-by-side with the original for easy comparison.
4. Includes basic options to:
• toggle between “standard” and “simplified technical” output modes,
• set creativity/temperature, and
• copy results to clipboard.
Tech stack is flexible; Python + Flask, Node + Express, or a lightweight desktop framework are all fine as long as setup is straightforward on Windows or Linux. Package dependencies, installation steps, and an API key placeholder must be documented in a concise README.
Deliverables
• Source code with clear comments.
• README covering local setup, model/API configuration, and usage instructions.
• A brief demo video or screenshots showing the prototype in action.
If this initial build works well, I’ll expand the project to include batch processing and CMS integrations. I would like to use this as a webpage and a iPhone applicwtion.
Core objectives
• Primary function: improve text to sound more human.
• When the input is a technical document, the output must read in a clearly simplified tone that cant be detected by ai.
• For all other text types, the tool should keep the original intent while adding natural flow, contractions, and everyday vocabulary.
I would also like a % ai use
Scope for this phase
I’m looking for a working prototype that:
1. Accepts plain-text input (web form or simple desktop interface—your choice).
2. Uses an LLM (OpenAI, Hugging Face, or similar) with prompt engineering or fine-tuning to perform the humanization.
3. Returns the rewritten text instantly and displays it side-by-side with the original for easy comparison.
4. Includes basic options to:
• toggle between “standard” and “simplified technical” output modes,
• set creativity/temperature, and
• copy results to clipboard.
Tech stack is flexible; Python + Flask, Node + Express, or a lightweight desktop framework are all fine as long as setup is straightforward on Windows or Linux. Package dependencies, installation steps, and an API key placeholder must be documented in a concise README.
Deliverables
• Source code with clear comments.
• README covering local setup, model/API configuration, and usage instructions.
• A brief demo video or screenshots showing the prototype in action.
If this initial build works well, I’ll expand the project to include batch processing and CMS integrations. I would like to use this as a webpage and a iPhone applicwtion.