Knowledge Graphs using LLMs
Structuring unstructured data using OCR, LLMs, and Knowledge Graph techniques.
Extracted and structured information from unstructured documents using OCR and Large Language Models (LLMs).
- Preprocessed 1000+ unstructured recipe documents and PDFs using OCR to extract raw text.
- Developed Python scripts to parse and standardize ingredient and recipe data.
- Converted structured data into knowledge graphs using LLMs (GPT-4, Mistral).
- Designed and optimized ETL pipelines to automate data flow, reducing processing time by 40%.
- Explored graph-based representations to enhance information retrieval and semantic search.