Skip to content
AK
Back to all projects

LLM and Agentic AI

Research prototype

Content Moderation AI

Problem

Harmful content appears across text, image, and video at the same time.

System

A multimodal moderation workflow that scores text and imagery with CLIP-based representations and routes uncertain cases for human review.

My contribution

Implemented the multimodal scoring pipeline and the human-in-the-loop routing logic.

System architecture

  1. Content input
  2. Multimodal encoding
  3. Classification
  4. Review routing

Technical focus

  • Multimodal AI
  • Content safety
  • Human-in-the-loop workflows

Technology stack

  • PyTorch
  • CLIP
  • Python

Data source

Dataset and data-collection write-up is being prepared.

Data pipeline

Ingestion and preprocessing write-up is being prepared.

Model or LLM workflow

Model and workflow write-up is being prepared.

Evaluation and results

Evaluation setup and results write-up is being prepared.

Challenges and engineering decisions

Key engineering decisions write-up is being prepared.