Project Overview
An AI system that classifies websites into IAB categories for digital marketing and ad-tech use cases.
The Problem
Accurate content classification at scale is difficult with rule-based systems.
The Solution
Used fine-tuned language models and prompt optimization to classify content reliably.
Architecture
- FastAPI inference service
- Prompt-tuned classification logic
- Taxonomy-driven output
Key Challenges
- Ambiguous content
- Category overlap
- Consistency across domains
Outcome
Delivered scalable and accurate content classification aligned with IAB standards.