Contact person: Business Client

Phone:Show

Email:Show

Location: Ludhiana, India

Budget: Recommended by industry experts

Time to start: As soon as possible

Project description:
"Build a Product Recommendation Web App that uses AI to recommend furniture products
and generate creative product descriptions. The application should be developed using
FastAPI (backend) and React (frontend), integrating with a vector database. ( dataset will be provided)
- The recommendation page will be a like back and forth conversation page.
- The analytics page will be on a different route with analytics of all the items in the
database.
Project Requirements
1. Machine Learning (ML): Build a product recommendation model based on the given
dataset. You may use any techniques you may find fit for this.
2. Natural Language Processing (NLP): Implement text analysis features to process
products and group similar or related products.
3. Computer Vision (CV): Develop an image classification model that can identify product
categories/types from the provided product images.
4. Generative AI(GenAI): Use a lightweight GenAI model(your choice)to generate creative
product design descriptions to accompany recommended products on the frontend.
5. Vector Database: Store embeddings (text/image-based) in a vector database for
semantic search and retrieval. You can use any vector database of your choice (Pinecone
preferred).
6. Frontend (React): Create a simple yet functional frontend to display product
recommendations once someone sends a prompt. These recommendations should come
with generated descriptions and product images..
7. Analytics Page: Add a page (using React routing) showing analytics on the current
dataset. You may use a different database for analytics if you wish.
Tech Stack Requirements
- Backend: FastAPI
- Frontend: React(any library/framework of your choice for UI)
- VectorDB: Pinecone (preferred) or any other vector DB
- ML Models: Any framework (e.g., scikit-learn, PyTorch, TensorFlow)
- NLP: Any model/tool (e.g., spaCy, HuggingFace Transformers)
- CV: CNN/ResNet/Vision Transformer/GenAI model (your choice)
- GenAI: Any open-source or API-based lightweight model
- Integration Framework: LangChain (must be used for GenAI or embedding-based tasks)
Deliverables:
You must submit a GitHub repository containing:
1. Frontend (React App)
2. Backend (FastAPI App)
3. Data Analytics Notebook (.ipynb) — must include:
- Clear and descriptive comments explaining your reasoning for each major step
4. Model Training Notebook (.ipynb) — must include:
- Clear and descriptive comments explaining your reasoning for each major step
- Model performance evaluation (if applicable)
5. Instructions ([login to view URL]) — detailing setup, usage, and environment requirements." (client-provided description)