A Partnership for Responsible AI: Truefoundry and Enkrypt AI
Imagine the power of giving every doctor and patient a smart assistant that understands complex medical information. Generative AI makes this possible, but a single wrong answer can have serious consequences. This is the challenge every innovative company faces: how to use this powerful technology responsibly and safely?
To solve this, TrueFoundry and Enkrypt AI are proud to announce a partnership. We're bringing together our expertise to offer a complete solution for the governance, control, and compliance of Generative AI. Our goal is to give every organization’s leadership and security teams the visibility and control they need to innovate with confidence.
“The combination of TrueFoundry’s fast, flexible ML deployment and Enkrypt’s security and compliance layer gives us exactly what we need to safely scale GenAI across our regulated workflows. We can now move from prototype to production with confidence, without compromising on patient safety, data privacy, or audit readiness”, - Senior PM, AI & Data Platforms, at a Global Medical Device Company.
The TrueFoundry AI Gateway: Your Central Control Center
Think of the TrueFoundry Gateway as the central hub for all your AI models. It provides the solid foundation you need for operations and monitoring, allowing you to manage everything from one place.
Key features include:
Easy management of different models from various providers.
- Detailed user management to control who has access across your teams.
- Powerful and clear insights with logs, metrics, and usage data.
- The ability to plug in custom security tools, like guardrails, directly into the gateway.
Enkrypt AI: Your AI Security Shield
Enkrypt AI acts as the intelligent security layer, protecting your AI from risks. They provide practical, fast, and highly accurate guardrails that are essential for preventing issues and maintaining compliance.
Key features include:
- Advanced guardrails to reduce issues like bias, toxicity, and incorrect information (hallucinations).
- Detection and prevention of actions that violate company policy.
- Specialized security tests, known as red teaming, to ensure your AI is safe for regulated industries like finance and healthcare.
Integration Guide: How TrueFoundry and Enkrypt AI Work Together
Allow our documentation to configure and deploy the integration:
TrueFoundry + Enkrypt AI Integration Guide
The integration between TrueFoundry’s AI Gateway and Enkrypt AI gives organizations a unified way to manage, secure, and monitor LLMs in production—especially in regulated environments like healthcare and finance.
How It Works
- Route Through the Gateway
Connect your AI applications to the TrueFoundry Gateway for visibility, access control, and observability. - Add Enkrypt Guardrails
Plug Enkrypt AI into the Gateway to monitor inputs and outputs in real time, block unsafe completions, and enforce policy-based controls. - Run Red Teaming ( Optional)
Enkrypt can simulate adversarial prompts to stress-test your system pre-deployment. - Monitor and Audit
All activity is logged within the Gateway, enabling clear traceability and reporting.
Putting It to the Test: A Healthcare Company's Journey
Let's look at a real-world example. A leading healthcare technology company created an AI-powered virtual assistant to help patients and technicians use their diagnostic equipment. 11While a great idea, they needed to ensure it was fully compliant and accurate.
The Hurdles They Faced:
The company noticed some serious problems.
- The assistant gave inconsistent advice about important procedures like device calibration.
- It sometimes invented incorrect information (hallucinations) when asked unusual questions.
- This exposed the company to major regulatory risks under rules like HIPAA and FDA guidelines.
- The team had no way of seeing when or how these risky interactions were happening.
A Three-Step Solution:
- Gaining Control with the Gateway: First, the chatbot was connected to the TrueFoundry Gateway. This instantly gave the team visibility into all requests and responses and allowed them to set up user-level access controls.
- Adding Guardrails for Safety: Next, Enkrypt AI’s guardrails were plugged into the gateway. These were set up specifically to catch medical misinformation and incorrect dosage advice, with instant alerts for any violations.
- Finally, Enkrypt AI’s automated red teaming put the system to the test. It sent tricky and harmful questions to the chatbot, such as how to misuse a device, to see how it would respond.
The Transformative Results:
The improvements were dramatic and measurable. The company achieved:
- A 70% decrease in hallucinations.
- 90% compliance with FDA policy tests.
- A 95% reduction in unsafe or incorrect outputs.
- 100% visibility into how the tool was being used across the company.
- A much faster cycle for testing, deploying, and fixing the model.
- Most importantly, they passed their external compliance audit using the reports from Enkrypt AI.
Who is This For?
This combined solution is perfect for:
- Healthcare and other regulated industries that need to meet strict compliance rules.
- Product and Platform teams who want to launch new features safely without ruining the user experience.
- Security and Compliance officers who need reliable, proactive tools to manage AI risk.
Learn More
- To dive deeper into the TrueFoundry AI Gateway and its capabilities, check out the official TrueFoundry Documentation.
- To discover more about Enkrypt AI's suite of security and guardrail solutions, visit the Enkrypt AI Documentation.
Frequently Asked Questions
AI governance is the framework of policies, controls, and oversight that ensure generative AI systems operate safely, compliantly, and within organizational risk tolerance—critical in healthcare and finance where errors have direct consequences. Learn more about AI governance terms.
- Prevents unsafe outputs like hallucinations and biased recommendations in patient care
- Ensures audit readiness and compliance with regulatory requirements
- Centralizes visibility and control across all AI deployments and models
Secure generative AI in production by routing all AI applications through a central gateway with real-time policy-based guardrails that monitor inputs and outputs, block unsafe completions, and enforce compliance controls. Enkrypt AI's runtime guardrails reduce manual compliance effort by up to 90%.
- Deploy guardrails at the gateway to catch bias, toxicity, and hallucinations instantly
- Enforce company policies automatically without slowing model inference
- Maintain detailed logs and metrics for audit and incident response
Compliance monitoring tracks live AI behavior against policies in real time; red teaming proactively simulates adversarial attacks and edge cases to stress-test systems before they reach production. Both are essential—monitoring catches violations, red teaming prevents them.
- Monitoring detects policy violations and unsafe outputs during normal operation
- Red teaming identifies hidden vulnerabilities across 300+ risk categories
- Together they provide defense-in-depth for regulated AI deployments
TrueFoundry and Enkrypt AI together provide the complete solution: TrueFoundry's AI Gateway offers centralized model management and observability, while Enkrypt AI adds intelligent security guardrails and specialized red teaming for life sciences compliance.
- Move from prototype to production without compromising patient safety or data privacy
- Achieve audit readiness with policy-based controls and detailed compliance logs
- Stress-test AI systems with red teaming designed for regulated healthcare environments
Enkrypt AI's guardrails catch hallucinations and policy violations before they reach production. Book a demo to see how it protects your regulated workflows, or start a free trial to test it yourself.
.jpg)
.svg){kind=small}
