jayesh0012/text-integrity-studio

Text Integrity Studio

A web app with:

• —Frontend text utility interface
• —Node backend API
• —Private model-service scaffold for your own trained humanizer model
• —Local plagiarism-style similarity scanner

Run The Web App

npm start

Open:

http://localhost:4173

Connect Your Own Model

The web backend calls your private model through MODEL_ENDPOINT.

Example:

$env:MODEL_ENDPOINT="http://localhost:8080/humanize"
npm start

When deployed, set MODEL_ENDPOINT to your hosted model-service URL.

Train Your Model

Put training pairs in model-service/dataset.jsonl:

{"input":"AI-looking text here","output":"Your preferred human-written version here"}

Install model dependencies in a Python GPU environment, then run:

cd model-service
pip install -r requirements.txt
python train_lora.py --dataset dataset.jsonl --output-dir trained-humanizer

If you have AI-vs-human CSV files, rebuild the prepared datasets:

cd model-service
node generate_dataset.js
node prepare_real_datasets.js "C:\Users\ASUS\Downloads\ai_human_detection_v1.csv" "C:\Users\ASUS\Downloads\ai_vs_human_text.csv"

Then serve it:

uvicorn serve:app --host 0.0.0.0 --port 8080

Google Deployment Shape

Deploy two services:

• —text-integrity-studio: Node web app
• —model-service: Python model API, preferably on GPU infrastructure

Set this environment variable on the Node web app:

MODEL_ENDPOINT=https://YOUR-MODEL-SERVICE-URL/humanize

Set the public website URL too, so canonical tags and the sitemap use your real domain:

PUBLIC_SITE_URL=https://yourdomain.com

SEO Setup

The app includes:

• —SEO title and meta description on the home page
• —Structured data for the web writing app
• —Crawlable content pages for writing and originality topics
• —robots.txt
• —sitemap.xml
• —About, Contact, Privacy Policy, and Terms pages

After deployment:

1. 1.Replace support@example.com with your real business email.
2. 2.Set PUBLIC_SITE_URL to your real domain.
3. 3.Submit https://yourdomain.com/sitemap.xml in Google Search Console.
4. 4.Add real articles based on actual user questions, not copied or mass-produced text.
5. 5.Add Google AdSense only after the site has useful content and a privacy policy.

Commercial Readiness

See COMMERCIAL_READINESS.md for the current launch status, model quality gates, and what still needs to happen before public monetization.