EU AI Act enforcement in

Prove what happened to any file.
To any auditor. At any point in time.

EU AI Act penalties start at 7% of your annual revenue. For a $10 million company, that is $700,000. Signal Provenance is $10,000 a year. Two minutes to set up.

49

Initiatives completed

1,085

Actions tracked with provenance

615

Verified completions

We ran it on ourselves first
real data from ops.db
3,473

documents scanned

18

structural gaps found

82%

time saved

122

initiatives tracked

8 tools, no single view. Half the operational knowledge in one person's head.

Same system deployed for clients.

See the full build log

Signal

We show you what your business actually looks like.

Built from the information you already have. What exists, what depends on what, where the gaps are, what to do next. Not a report that expires. A living model that gets smarter over time.

Signal Audit, $25,000

Full operational scan from your existing documents. Structured model of what exists, what depends on what, and where gaps are. One week.

Replaces: $50K-$150K consulting engagement

Signal Build, $50,000

A system that knows what your best people know. Living operational model on your hardware. Gap detection, dependency mapping, scored opportunities.

Replaces: $150K-$500K digital transformation

Signal Operate, $7,500/mo

It gets smarter over time. Weekly intelligence, monthly re-optimization, continuous document processing, automated violation detection.

Replaces: $8K-$25K/month fractional COO

EU AI Act, August 2, 2026

If your business uses AI tools, you have compliance obligations starting August 2.

The EU AI Act applies to US companies when AI output is used in the EU, broader than GDPR. Signal Provenance provides verifiable file provenance for $10,000 a year. Works wherever your files live. Two minutes to set up.

Get Provenance, $10,000/yr

What Signal finds

Real findings. Real numbers.

The company that built it

164 initiatives, 1085 steps, 5 engines validated, 82% time savings. This is not a demo. This is our actual operating infrastructure.

See the build log →
Engineering document set (internal)

Signal processed 3,473 engineering files from our own document corpus. Found 6 structural gaps across project types. Signal Audit tier: $25,000.

Multi-facility operations

A distribution company with operations across multiple facilities. Signal is building their operational model from internal documents: SOPs, contracts, compliance records. First client engagement, in progress.

First deployment

The system deployed on itself.

43 findings across 4 categories. Two human review gates. 100% confirmation rate. Every number from ops.db.

First Deployment -- echology-002
Real data from ops.db. The system deployed on itself.
43
Findings
98%
Resolved
2
Human Gates
5
Change Proposals
What Signal found
Scattered Knowledge14 findings
Process Gap12 findings (4 critical)
Undocumented Dependency10 findings
Key Person Dependency7 findings (4 critical)
Human review gates
Raw Gate
22/22 confirmed
100%
Optimized Gate
21/21 confirmed
100%
42/43 findings resolved
Every finding traced to source. Every resolution tracked. Nothing ships without approval.

Source: ops.db, deployment echology-002

Your data stays on your machine. Your proof travels with you.

Works wherever your files live

Local folders, Google Drive, SharePoint, S3. The chain holds regardless. Your cloud provider cannot rewrite it. No one can.

Nothing ships without your approval

Every finding is reviewed before it reaches you. Every output is traceable to its source. You see exactly where each conclusion came from.

You keep everything

No lock-in. No subscription to maintain access. The system and the model it builds are yours.

Prove any file's integrity. Satisfy provenance requirements. Backed by our refund guarantee.

Signal Provenance. $10,000 a year. Two minutes to set up. If your organization is fined for a provenance failure that Signal was configured to track, we refund your subscription in full.

Kyle Vines

Kyle Vines

Founder

I acquired a geotechnical firm and watched what happens when key people leave. Half the operational knowledge walked out the door. I modernized every legacy system in that company, but the real problem was upstream: critical information was scattered across systems, people's heads, and filing cabinets. No tool could reason about it because nobody had organized it first.

At LaBella Associates (ENR Top 200), I deployed AI tools across a 200-person engineering team. Same pattern. The tools that worked had structured information underneath them. The ones that failed were thrown at raw documents. Signal is the structural step that was always missing: perceive what exists, classify it, present it with provenance. Then the reasoning works.

Acquired and ran a geotech firm 13+ yrs civil/geotech LaBella Associates (ENR Top 200) AI deployed across 200 engineers