Gram_Sarthi

Gram Sarthi addresses a critical accessibility gap: rural Indians struggle to navigate complex legal rights, healthcare guidance, and government scheme eligibility due to language barriers. To solve this, a bilingual RAG pipeline was built on Databricks that ingests 39,000+ records across legal acts, medical Q&A, and government schemes, processes them through a Medallion architecture into 51K searchable Hindi-translated chunks, and serves accurate answers in simple Hindi.

Built With

• pyspark
• python
• sql