HX-Provenance - Verifiable Evidence and Provenance Appliance

Overview

HX-Provenance is a customer-controlled provenance appliance that turns important digital artifacts into verifiable evidence. Running as a CPU-only VM inside your AWS environment, it issues cryptographically signed provenance receipts and exportable evidence bundles - giving regulated, enterprise, and public-sector teams a repeatable proof path for audit, compliance, legal, cyber, AI governance, and critical decision workflows.

Unlike SaaS provenance services that hold signing keys externally or blockchain-based audit trails with latency and cost trade-offs, HX-Provenance keeps the entire operating boundary - keys, data, and verification logic - under your direct control.

Key Capabilities

Signed Provenance Receipts: Issue ML-DSA-65 (post-quantum) signed receipts for registered artifacts including AI outputs, retrieval results, compliance files, legal packets, cyber incident artifacts, procurement materials, board packets, case files, and other critical decision documents.

Evidence Bundle Export: Package receipts, metadata, and verification material into self-describing evidence bundles for legal handoff, audit review, and chain-of-custody workflows.

Online and Offline Verification: Verify receipts through the appliance API or offline using exported verification material - no always-on connectivity required.

100% Artifact Recall: Every registered artifact is retrievable and verifiable, eliminating gaps in audit trails and chain-of-custody records.

Performance

In benchmark testing, HX-Provenance produced signed verification receipts in approximately 240 ms end to end on a cloud CPU instance, achieving approximately 4.17 receipts per second. This gives teams a fast, repeatable proof path without GPU dependencies.

Deployment Requirements

Instance: CPU-only VM. Recommended baseline: 4 vCPUs, 16 GB RAM.

Storage: Minimum 10 GB boot storage, 100 GB persistent data storage for production.

IAM Permissions: Deploy and operate the VM, manage network and firewall access, access serial console or startup credentials as needed.

Optional AWS Integrations: Write evidence bundles to customer-controlled Amazon S3 buckets. Send telemetry to customer-controlled Amazon CloudWatch.

Network: Customer-managed VPC with security group rules for HTTPS/API access.

Integration Points

HX-Provenance integrates with internal applications, AI/ML pipelines, CI/CD systems, compliance tooling, records systems, and evidence workflows through its REST API. Customers can optionally connect the appliance to Amazon S3 for evidence bundle storage and Amazon CloudWatch for operational telemetry, keeping all data within their AWS account.

Security and Cryptography

Signing Algorithm: ML-DSA-65 (post-quantum digital signature).

Deployment Isolation: Runs entirely within the customer's VPC with no external call-home or key escrow.

Key Control: All signing keys remain under customer control within the appliance.

Data Boundary: Artifacts, receipts, and evidence bundles never leave the customer's AWS environment unless explicitly exported.

Use-Case Scenario

A compliance team at a financial institution uses HX-Provenance to register AI-generated risk assessment reports. Each time the ML pipeline produces an output, the appliance issues a signed receipt binding the report content, model version, input data hash, and timestamp into a verifiable record. When auditors request evidence months later, the team exports an offline-verifiable evidence bundle - proving exactly what the model produced, when, and from what inputs - without relying on the appliance being online at audit time.

Getting Started

To evaluate HX-Provenance, launch the AMI on a supported instance (4 vCPUs, 16 GB RAM minimum), complete initial HTTPS and API key configuration, and issue your first signed receipt. For guided evaluation or pilot deployment planning, contact HolonomiX through the support channel.

HolonomiX is the developer of HX-Provenance, focused on verifiable provenance and evidence integrity for organizations that need to prove what happened across AI, compliance, legal, and decision workflows.