We use cutting-edge AI.
Your data is never used to train it.
At FoundryOps, you don't have to choose between powerful AI and data privacy. We built our platform to deliver both.
Our Privacy Architecture
Enterprise AI, Zero Training
We use Google Vertex AI within our GCP project under enterprise data protection terms. Your data is never used to train AI models—by us or by Google.
PII is masked before AI processing. Emails become ***@domain.com.
Your Data Isn't Our Product
We make money by solving your data quality problems, not by selling your data. We will never:
- Resell or share your customer data
- Use your data to train models for other customers
- Allow your records to improve anyone else's matching
- Create "data co-ops" or "collaborative intelligence" from your CRM
Enterprise Security
- TLS encryption in transit
- JWT authentication with role-based access
- Isolated customer workspaces
- GDPR-ready data handling
How We Use AI (The Right Way)
✓ What Our AI Does
- Uses Google Vertex AI under enterprise data protection
- Processes in isolated workspaces per customer
- Operates in inference-only mode (no training on your data)
- Masks PII before AI processing (emails → ***@domain)
✗ What Our AI Doesn't Do
- Send your data to ChatGPT/Claude/Gemini
- Train on one customer's data to benefit another
- Store or learn from your records after processing
- Share anything with model providers
Built on Open Data, Not Your Data
Our matching intelligence comes from the Foundry Graph—a knowledge base built from public sources like Wikidata, SEC EDGAR, and Companies House. Not from harvesting your customer records.
This means our accuracy improves from better algorithms and open data—never from learning on your customer records.
Your competitor will never benefit from your CRM data.
Compare Our Approach
| Privacy Practice | FoundryOps | Typical Vendors |
|---|---|---|
| AI Model Location | Enterprise-grade, no training (Google Vertex AI + DPA) | External APIs (OpenAI, Anthropic, etc.) |
| Training on Your Data | Never | "Improves our models" |
| Data Sharing | Zero. Isolated workspaces. | "Collaborative intelligence" |
| Business Model | Subscription for features | Data resale & insights |
| Data Ownership | Export & delete anytime | "Up to 90 days" |
"Because if you can't audit your AI, you don't really control it."
Our Promise to You
We'll tell you exactly how we handle your data. No buried terms, no "legitimate interest" loopholes. This page is proof of that commitment.
We charge for matching accuracy and workflow automation—not for aggregating and reselling your information.
Export everything anytime. Delete on request. No vendor lock-in through data hostage-taking.
Every feature we build starts with the question: "How do we do this without compromising customer privacy?"
Frequently Asked Questions
Q: If you use AI, how can you guarantee privacy?
A: We use Google Vertex AI under their enterprise Data Processing Addendum—your data is never used to train models. We also mask PII before AI processing (emails become ***@domain.com) and maintain isolated workspaces per customer.
Q: Do you use customer data to improve your matching?
A: No. Our matching improves through better algorithms and our open-source Foundry Graph (built from Wikidata and public registries), not by learning from customer data. Your competitor will never benefit from your CRM records.
Q: Can employees see our data?
A: Only with explicit permission for support purposes, under strict access controls with full audit logging. Customer workspaces are isolated, and access is role-based with JWT authentication.
Q: What about data deletion?
A: You can export or delete your data at any time. We don't hold your data hostage or impose arbitrary waiting periods. When you delete, it's gone—not archived for "compliance reasons."
We're an AI company that respects boundaries.
In a world where every vendor wants to train on your data, share your insights, and monetize your information, we take a different path: We sell great software, not your data.
Your trust is our business model.