I build small, focused software utilities designed to do one job well and return predictable machine-readable results.

Current work focuses on:

• Python automation
• AI-agent utilities
• Document processing
• Structured data extraction
• RAG and retrieval tooling
• APIs and machine-to-machine workflows
• Pay-per-use software primitives

The goal is simple: build useful machines, publish them, measure real usage, and improve the tools that prove their value.

Tools should be:

• Easy to understand
• Easy to call programmatically
• Predictable in their output
• Useful without manual onboarding
• Resilient to partial failures
• Transparent about pricing
• Composable with larger automated systems

PDF Citation Chunker

Converts public PDFs into page-aware, overlapping JSON chunks with source provenance, hashes, token estimates, batch processing, failure isolation, and machine-readable run summaries.

More utilities are being developed through the EchoEmpire Machine Foundry.