Independent systems researcher and builder based in Seville, Spain.
I build infrastructure for verifiable and resumable AI workflows. My current project is HREVN, a public-alpha integrity layer for signed checkpoints, structured evidence, and audit-ready workflow continuity.
- durable and resumable workflow state
- tamper-evident evidence and provenance
- reconciliation across software and operational boundaries
- MCP and agent-tool integration
- explicit separation between executed evidence, source analysis, and architectural inference
hrevn-workflow— local-first Python SDK for checkpoints, resume, artifact hashing, and workflow manifestshrevn-mcp-server— thin MCP surface for the managed HREVN runtimerwa-collateral-reconciliation— experimental reference model for condition-evidence provenance, versioned valuation attestations, and reconciliation with on-chain collateral state
I am studying the continuity and reconciliation problems that appear when a working proof of concept must become an operational system: durable state, retries, finality, external ledgers, evidence provenance, and recovery after partial execution.
- Web: hrevn.com
- Email: contact@hrevn.com
Public repositories are labeled according to their maturity. Experimental work is kept private until it has a reproducible public path.
