Smart Product Pricing Ensemble Model

Our solution for Amazon ML Challenge 2025 that ranked 52 out of 9,000+ teams.

Achieved rank 52 out of 9,000+ teams (Top 0.6%) in Amazon ML Challenge 2025 with a multimodal architecture for product price prediction.

  • Dual-stream pipeline using DeBERTa, ELECTRA, and CLIP-ViT with attention-based fusion and residual blocks.

Tech Stack

  • Python, Hugging Face, PyTorch, Pandas, NumPy, NLTK, scikit-learn