AI Transformation Unleashed: Stories from Silk Customers and Microsoft Experts
Topic: Real-World AI Innovation Across Healthcare, Engineering, and Data Platforms
Speakers:
Nicholas Clark, AI Strategist, Silk
Kellyn Gorman, Data & AI Specialist, Silk
Eduardo Kasner, Chief Data & AI Officer, Microsoft
Steve Dertien, EVP & CTO, PTC
Jeff Thomas, CTO & SVP, Sentara Health
Jon Copley, Chief Innovation & Digital Technology Officer, Franciscan Health
Summary
This session brings together AI strategists, Microsoft leaders, and Silk customers to explore how Silk’s AI enablement accelerates data-driven transformation across industries.
The discussion centers on real customer use cases in healthcare and engineering, showing how organizations leverage Silk’s technology with Microsoft’s Azure AI Foundry and SQL 2025 to deliver scalable, secure, and high-performance AI applications.
Key Takeaways
Data is the fuel for AI, and Silk ensures it’s accessible, performant, and secure.
Azure AI Foundry provides a unified framework for managing multiple models and agents.
Silk’s cloning and snapshot technology reduce latency and enable real-time AI workloads safely.
Customers like PTC, Sentara, and Franciscan Health use Silk to balance performance with governance and scale AI responsibly.
The future of AI is agentic, trustworthy, and data-driven — built on foundations of zero trust and intelligent data management.
Transcript
[00:00 – 02:00] Welcome and Agenda
Nicholas Clark (Silk):
Welcome, everyone! I’m Nicholas Clark, an AI strategist with over 25 years in data and AI, including my time at Microsoft. Today I’m thrilled to represent Silk as we highlight how Silk and Microsoft together empower organizations to unlock the full potential of their data.
I’m joined by my partner in crime, Kellyn Gorman — a Data and AI Specialist and multi-platform DBA with decades of experience at Microsoft, Oracle, and leading startups.
Here’s what to expect:
A keynote from Eduardo Kasner, Microsoft’s Chief Data & AI Officer.
Customer stories from PTC, Sentara Health, and Franciscan Health.
A roundtable discussion on shared challenges, lessons, and AI success patterns.
We’ll explore how Silk and Microsoft address scalability, governance, and performance to enable responsible, production-grade AI.
[02:00 – 09:00] Keynote – Eduardo Kasner (Microsoft)
Topic: Making AI and Your Data Real
Eduardo Kasner (Microsoft):
It’s a pleasure to be here. Let’s talk about how AI has evolved — from simple chatbots to multi-model intelligent applications combining text, audio, and vision.
With this shift come new challenges:
Data quality and accuracy
Security and compliance
Continuous improvement and observability
To support this, Microsoft launched Azure AI Foundry, a comprehensive platform integrating OpenAI, Hugging Face, NVIDIA, Meta, Databricks, and others — giving customers full control over governance, safety, and deployment.
Foundry provides:
A catalog of foundation and industry-specific models
1,400+ Logic App connectors
Secure BYO storage and VNet support
Full integration with Copilot Studio, Visual Studio, and GitHub
Kasner emphasized that AI’s fuel is data, and that data often lives in production systems — creating risk for performance and security.
That’s where Silk comes in.
Silk’s real-time cloning, zero-latency snapshots, and secure multi-environment deployment prevent production slowdowns, ensure compliance, and accelerate AI development.
In benchmarks, Silk delivered up to 6.5× faster semantic search performance compared to native Azure disks — proving its power in AI-driven workloads.
[09:00 – 15:00] Customer Story #1 – Steve Dertien (PTC)
Topic: Engineering Innovation with AI and Data
Steve Dertien (PTC):
PTC builds software for engineers — powering industries from aviation and automotive to medtech and industrial design.
Our journey with Silk began as we shifted from on-premises software to SaaS delivery. Managing hundreds of customer datasets and large-scale engineering data meant we needed a performant, scalable, and cost-efficient backend.
Silk gave us:
Efficient data management for thousands of workloads
Improved operational control for multi-tenant systems
Performance isolation — preventing AI inference or retrieval workloads from impacting production
As we integrate AI agents and graph-based data structures, Silk helps us organize complex relational data into more flexible, AI-ready formats — without rewriting our legacy applications.
[15:00 – 27:00] Customer Story #2 – Jeff Thomas (Sentara Health)
Topic: Building Secure Data Enclaves for Healthcare AI
Jeff Thomas (Sentara Health):
Sentara Health is a century-old integrated health system with 12 hospitals across Virginia and North Carolina. We manage both a health plan and a health system — meaning massive volumes of highly sensitive data.
Our cloud journey began six years ago to increase scalability, security, and innovation speed. Using zero-trust architecture, we moved to Azure and started building an AI landing zone — with Silk playing a crucial role in ensuring secure, near real-time data access.
Generative AI demands fresh, relevant data — not yesterday’s batch. But real-time access can’t compromise performance.
That’s where Silk’s data enclave model helps:
We maintain real-time, read-only copies of production data for AI workloads.
Third-party models can access data inside our environment, without the data ever leaving our control.
This approach enables secure AI experimentation and cost-efficient scalability.
Silk’s cloning and snapshotting allow us to run production-grade AI workloads without impacting Epic or claims system performance. It’s a game-changer for healthcare.
[27:00 – 36:00] Customer Story #3 – Jon Copley (Franciscan Health)
Topic: Healthy Data, Healthy AI**
Jon Copley (Franciscan Health):
Following Sentara’s example, we migrated Epic to Azure with Silk and Microsoft. This enabled us to modernize our infrastructure and build AI systems that improve patient care and clinician efficiency.
Franciscan Health operates 11 hospitals across two states, serving both urban and rural populations.
AI helps us:
Address provider shortages and clinician burnout
Deliver real-time insights for doctors and nurses
Enhance patient engagement through secure digital touchpoints
Silk gives us velocity — real-time data flow across enclaves — and security — ensuring sensitive PHI stays protected.
We call it “organic AI” — clean, reliable, “healthy” data that fuels trustworthy outcomes.
Silk’s platform not only protects against cyber threats but also empowers us to scale responsibly. We can now focus on improving outcomes and restoring the human element of healthcare.
[36:00 – 54:00] Roundtable Discussion
Nicholas Clark (Moderator):
Let’s unpack common themes: data enclaves, agent-to-agent collaboration, and the role of Silk in zero-trust architectures.
Steve Dertien (PTC):
AI in engineering is a lot like healthcare data — it’s sensitive, regulated, and business-critical. Silk allows us to support isolated, multi-tenant environments while maintaining control over intellectual property.
Jeff Thomas (Sentara):
Our vendors used to ask for full database exports — which was unacceptable. Now, Silk gives us instantaneous snapshots for dev/test and secure production environments.
We’re also excited about Silk’s inline data masking capabilities with partners like Redgate — ensuring PHI remains protected across environments.
Jon Copley (Franciscan):
This architecture eliminates the old “nightly batch job” model. Our data remains live, accessible, and compliant.
Silk’s real-time cloning and throughput give us performance without compromise — vital for EHR, analytics, and AI workloads.
Jeff:
Exactly. AI’s effectiveness depends on zero-latency, zero-trust environments. Silk provides that bridge — turning compliance from a blocker into an enabler.
Steve:
And for PTC, Silk’s elasticity helps us handle unpredictable workloads from thousands of customers — securely and efficiently.
[54:00 – 56:00] Closing Remarks
Kellyn Gorman (Silk):
It’s inspiring to see AI moving beyond text generation to real-world action — healthcare improvements, engineering acceleration, and operational efficiency.
Companies are realizing that successful AI isn’t just about models — it’s about architecture, governance, and trust.
Nicholas Clark (Silk):
Exactly. Silk and Microsoft together are empowering customers to scale responsible, high-performance AI.
Keep an eye on Silk’s blog for new content from Tom O’Neill on AI enablement, and visit silk.us for demos or consultations.
Thank you, everyone — and special thanks to Eduardo, Steve, Jeff, and Jon for sharing your insights. See you at the next Silk session!
