Kudra

Kudra is a web-based document intelligence platform for businesses that need to automate document processing and extract structured data from complex files.

Key capabilities include:

• OCR ingestion for native PDFs and scanned documents, including formats such as PDF, JPG, PNG, DOCX, and Excel
• Deterministic extraction of entities, relations, tables, charts, and images, designed for complex layouts like multi-page financial tables and nested hierarchies
• AI enrichment with LLMs for reasoning tasks such as summarization, classification, and analysis, plus AI agents that can answer questions over extracted data
• Visual grounding and provenance that maps each extracted field back to its bounding box in the source document with confidence scores for auditability

Pricing: Free and paid tiers with monthly subscriptions, Starter (Free, 100 pages/month, 1 user), Basic ($299/month, 1,000 pages/month), Growth ($599/month, 5,000 pages/month, up to 10 users), plus Enterprise (custom) and pay-as-you-go for extra documents.

What makes it notable is its combination of deterministic extraction and LLM-based enrichment with source-linked provenance, along with 50+ pre-built extraction templates and API access for integrating document workflows into internal systems.