HelloCheck: OCR & LLM Nutrition Data Pipeline
Internal Proprietary ToolingImageEntity Ner Classification
Developed an end-to-end AI pipeline for "HelloCheck," a system that converts unstructured grocery receipt images into structured nutritional data. I managed the full data lifecycle, including:
Data Collection & Annotation: Curated and labeled a dataset of 1,000+ receipts to train and validate the OCR and LLM-based extraction engine.
Named Entity Recognition (NER): Implemented logic for precise extraction of food items, prices, and quantities from complex layouts.
Quality Control: Built a validation loop using Python to ensure 98%+ accuracy in item classification and health scoring (A–F).
Project Links: > * App: https://hellocheck.app
Product Hunt: https://www.producthunt.com/products/hellocheck
Developer Updates: x.com/plus8bit
Developed an end-to-end AI pipeline for "HelloCheck," a system that converts unstructured grocery receipt images into structured nutritional data. I managed the full data lifecycle, including: Data Collection & Annotation: Curated and labeled a dataset of 1,000+ receipts to train and validate the OCR and LLM-based extraction engine. Named Entity Recognition (NER): Implemented logic for precise extraction of food items, prices, and quantities from complex layouts. Quality Control: Built a validation loop using Python to ensure 98%+ accuracy in item classification and health scoring (A–F). Project Links: > * App: https://hellocheck.app Product Hunt: https://www.producthunt.com/products/hellocheck Developer Updates: x.com/plus8bit
2025 - 2025