How to Use the Compliance & Moderation Bot

Our Compliance & Moderation Bot is designed to secure your user inputs and data pipelines with just a few clicks.

Whether you are screening a single text prompt or bulk-processing thousands of user-uploaded images, this guide will show you how to submit your data, choose your workflow, and understand the comprehensive safety reports you receive.

Step 1: Prepare Your Data

Our engine is highly flexible and accepts multiple data formats.

  • Text Inputs: Standard text strings, user chat logs, or GenAI prompts.
  • File Inputs: Standard image files or documents.
  • IP Addresses (Optional but Recommended): To utilize our advanced Geo-Compliance features, include the user's IP address so the bot can enforce regional laws.

Zero Data Retention Guarantee: We process your data with strict confidentiality. We do not save, log, or store your text inquiries, file attachments, or uploaded documents. The system reads your data securely in memory to generate the compliance score, and then immediately deletes it. Nothing is ever saved to our servers, and your data is never used to train outside AI models.

Step 2: Choose Your Submission Method

We support five main ways to submit your data, depending on your workflow size and automation needs:

  • Method 1: Direct Text Paste: Best for quick tests of specific user prompts or chat snippets.
  • Method 2: File Attachment: Best for one-off analyses of single images or documents.
  • Method 3: Cloud Sync (BYOS): Best for automated, bulk processing via Google Drive or Dropbox. We pull the files, process them, and deliver a master CSV report back to your cloud folder.
  • Method 4: Developer API: Best for large-scale enterprise automation, background processing, and real-time app integration.
  • Method 5: Custom Managed Services (Done-For-You): Best for teams who want our engineers to build their pipeline and customize the bot's rules (e.g., adding specific community guidelines, updating local laws, or creating custom risk categories for youth safety).

Review the comparison table below to choose the method that best fits your needs:

Features & Requirements Method 1 & 2: Direct Text & File Method 3: Cloud Sync (BYOS)
Input Format Paste text or upload a single file directly into the hub. Connect your Google Drive or Dropbox folder containing your files.
Processing Style Instant, real-time processing. Automated bulk background processing.
Output Delivery JSON compliance report returned instantly on-screen. Master CSV report automatically synced back to your Cloud Drive.
Email Notifications N/A Yes. Receive an automated email alert with job details when your cloud folder is populated.

Step 3: Understanding Your Results (The Output)

Unlike standard filters that just say "Blocked," our bot provides a highly detailed, explainable JSON response for every input.

1. The 10 Severity Categories (Scored 0.0 to 1.0)

The bot scores the input across ten distinct risk categories (unless you have requested custom categories, like "Age-Restricted," via our Managed Services). A score of 0.0 means no risk, while a score closer to 1.0 indicates severe violation.

  • Harassment, Hate Speech, Sexually Explicit, Dangerous, Civic Integrity, Self-Harm, Copyright.
  • Jailbreak: Did the user try to trick or hack the AI?
  • Base Location Laws: Does it violate the laws of your company's home country?
  • IP Location Laws: Does it violate the local laws of the user's physical location?

2. The Final Result

Based on customizable thresholds, the bot will return one of three definitive actions:

  • SAFE: The content is clean and cleared for use.
  • FLAGGED: The content is suspicious and requires human review (e.g., borderline severity).
  • BLOCKED: The content is highly toxic, illegal, or a jailbreak attempt and must be stopped immediately.

3. The Reasoning Summary (Explainable AI)

The most powerful feature of the output. The bot provides a plain-English explanation of its decision.

Example: "The input contains a targeted negative comment towards an individual. Result: BLOCKED."

Step 4: Run the Engine & Retrieve Your Data

Once you click submit, the engine will process your data based on your chosen method:

If using Method 1 or 2 (Direct Text / File):

The engine processes your data immediately. Within seconds, your structured JSON compliance report will be displayed directly on your screen, ready for your app to ingest.

If using Method 3 (Cloud Sync / BYOS):

Because cloud folders often contain thousands of files, this method automatically runs in the background. You can close the window and continue working. Once finished, you will receive a Job Completion Notification Email detailing the processing time and files scanned. Simply open your Google Drive or Dropbox to find your newly generated moderation_results.csv file containing the safety scores for every file.