Haawke Neural Technology

Provenance infrastructure for human–AI collaboration

Models can change. The record remains.

Haawke Neural Technology develops tools that preserve verifiable records of AI-assisted work—from generation and source material to edits, derivatives, and published artifacts.

Our focus is provenance: recording what was produced, which systems participated, and how artifacts relate across a creative or research workflow. We combine cryptographic fingerprints, verification records, and independent timestamping with practical tools for creators, developers, and researchers.

Core Focus Areas

• Cryptographic provenance: SHA-256 artifact fingerprints, registration receipts, public verification, and OpenTimestamps proofs anchored to Bitcoin.
• Human–AI contribution history: Preserving original human work, model outputs, and subsequent versions as distinct, connected records.
• Multimodal creative workflows: Provenance for text, images, music, and the source materials behind them.
• AI memory and continuity: Authenticated historical records that support collaboration across models, providers, and discontinuous sessions.
• Integration: API and Model Context Protocol interfaces for hashing, registering, and verifying artifacts.

Haawke Projects

Haawke Phoenix

A provenance-native AI workspace combining multiple models with conversation history, artifact receipts, provenance exports, and public verification.

Haawke Hash

Artifact fingerprinting and registration infrastructure for preserving evidence of digital work.

Haawke Verify

Public verification of registered artifact fingerprints and associated provenance records.

Memory Chain

A system for preserving tamper-evident human–AI collaboration records through hashing, registration, and independent timestamping.

Squeek

Music and creative workflow development connecting generative tools with preservation, provenance, and distribution.

Our Approach

Provenance provides evidence. Humans decide what that evidence means.

A cryptographic record can support artifact-integrity verification and document recorded history. It does not certify factual accuracy, independently establish legal authorship, or determine creative intent.

We build across model and provider boundaries so that the history of a collaboration can remain useful as its tools evolve.

Collaborate

We welcome collaboration around provenance-aware models, datasets, creative pipelines, research tools, and interoperable artifact records.

• Website: haawke.com
• Phoenix: phoenix.haawke.com
• GitHub: inoculate23
• Contact: craig@haawke.com

Founded by Craig Ellenwood · Point Roberts, Washington