Every prompt is a data transaction. Govern every one.
Your teams want to use your best data in AI. Compliance keeps saying no. In 45 minutes, see how banks, insurers and health systems keep PII and PHI out of RAG pipelines and AI agents, without slowing AI down.
Save your seat
Free. Can't make it live? Register and we'll send the recording.
You're in.
Your calendar invite is ready. We'll email your link the day before, and the recording right after.
Got a question for the panel? Reply to your confirmation email and we'll try to answer it live.
From the team behind the de-identification engine trusted by Boehringer Ingelheim, Zurich Insurance and MUFG Bank
Three things you can use the next morning.
No product tour disguised as a webinar. Practical answers to the questions privacy, security and data teams are asking right now.
Where sensitive data leaks in AI
RAG retrieval, agent memory, prompt logs, fine-tuning sets. The four places PII and PHI slip through, and how to close each one.
What “audit-ready” really means
The evidence auditors accept for HIPAA expert determination, GDPR, CPRA and PCI-DSS, and how to produce it without a six-month project.
Why context beats pattern matching
A live demo of what rules miss: the SSN without the word “SSN”, the card number spoken across two turns, the name that doesn't look like a name.
See what a rule would miss.
Press the button. This is the kind of messy, real-world text we'll de-identify live, on screen, with the entities explained.
- Press “De-identify it” …
No “SSN” keyword, numbers spoken as words, a card split across a sentence. Pattern matching struggles here. Context doesn't.
Built for regulated data, in production today.
45 minutes, minute by minute.
Bring your hardest question. We save the last ten minutes for yours.
Welcome and a quick poll
Where are you with AI governance? We'll share the room's answers live.
Where PII and PHI leak in AI
RAG pipelines, agents, logs and training data, with real examples from regulated teams.
Live demo: context vs. rules
De-identifying call transcripts, clinical notes and documents, entity by entity.
From “no” to “yes”: a customer story
How one team got compliance to approve AI on their most restricted data.
Your questions, answered
Open Q&A with the speakers, plus the three most asked questions from registration.
Built for the people who say yes (or no) to AI.
If your job is to protect sensitive data, or to put it to work, this session is for you.
Patricia Thaine
Founder, CAIO and Chairwoman, Limina
A customer data leader
Financial services, sharing what changed in production
Your host
Limina marketing, keeping it on time and on topic
Teams that stopped choosing between compliance and useful data.
Real words from Limina customers, as published on getlimina.ai.
“Our platform reduces response time from days to minutes, but we couldn't have launched without proper anonymization. It was critical for compliance.”Israel Gonzales Salas · Data Engineer, Boehringer Ingelheim
“After doing a side-by-side comparison, it quickly became clear to us that we couldn't go back to using something like AWS.”Sebastian Jimenez · Founder and CEO, Rilla
Good questions, short answers.
I can't make it live. Should I still register?
Yes. Everyone who registers gets the recording, the slides and the guide “De-identification for AI Training Data” the next day.
Is it technical?
It's built for both sides of the table. Privacy and compliance leaders get the “what and why”; data and engineering teams get the “how”, including a live demo.
Will it be a sales pitch?
No. You'll see the product in the demo, because that's the clearest way to show the problem. The rest is practical and vendor-neutral.
Can I bring my team?
Please do. Use “Copy link for a colleague” after you register, or forward your confirmation email.
Does my data leave my environment with Limina?
No. Limina deploys in your VPC or on-prem, with no third-party access and no outbound calls. We'll show the architecture in the session.
Save your seat. Bring your hardest question.
Thursday, November 19 · 11:00 AM ET · 45 minutes · free