As organizations embrace the transformative potential of AI, security and governance have become mission critical. That’s why PointGuard AI—a pioneer in securing AI and application ecosystems—is proud to announce its featured role at the upcoming Databricks Data+AI Summit (DAIS), taking place June 9–12 in San Francisco.

As a featured Databricks partner, PointGuard AI has been invited to present in a session titled Driving Secure AI Innovation, alongside Databricks and Obsidian Security. This session will highlight not only the urgency of protecting AI models and pipelines but also the powerful, real-world solutions now available to enterprises.

A Powerhouse Panel on AI Security

The panel brings together an elite group of security and AI leaders:

• Mali Gorantla, Chief Product Officer at PointGuard AI
• JD Braun, Security Architect at Databricks
• Alfredo Hickman, Chief Information Security Officer at Obsidian Security

Together, they will explore how leading organizations are operationalizing security for AI development through a new discipline known as AISecOps (also referred to as MLSecOps or LLMSecOps). Modeled after DevSecOps, this emerging best practice bakes security into the AI lifecycle from model creation to runtime.

Their shared message: You can’t afford to bolt security on after deployment. It must be part of the foundation.

Built on the Databricks AI Security Framework

At the center of this discussion is the Databricks AI Security Framework (DASF) 2.0—a comprehensive control framework launched by Databricks to help enterprises navigate AI-specific risks and compliance challenges. DASF 2.0 spans 64 controls covering 62 risk vectors, offering actionable guidance for AI model hardening, access control, data integrity, and threat mitigation.

PointGuard AI is proud to be one of the first technology partners to fully integrate DASF into its platform, offering customers a seamless and streamlined approach to AI security on Databricks. This deep integration translates the abstract requirements of DASF into practical workflows, analytics, and protections embedded directly in the development pipeline.

“By mapping security findings directly to DASF controls, PointGuard AI makes regulatory compliance and operational security achievable in fast-moving AI environments,” said JD Braun of Databricks.

Live Demo: The End-to-End Security Platform for Databricks AI

In a hands-on live demo, Mali Gorantla will walk the audience through the key capabilities of the PointGuard AI platform—purpose-built for the Databricks ecosystem. Attendees will see how the platform operationalizes DASF and secures the AI lifecycle across four critical pillars:

1. AI Discovery

Using distributed scanning powered by Databricks compute, PointGuard AI automatically inventories all AI assets—models, datasets, containers, and inference endpoints—across the enterprise. This capability ensures visibility into what AI systems exist, where they live, and how they’re configured.

2. AI Hardening (Security Posture Management)

PointGuard AI’s platform detects misconfigurations, access control gaps, model theft risks, and supply chain threats across MLOps pipelines. Built on top of MLflow and the Databricks Jobs API, it enforces secure practices throughout development and deployment.

3. Automated AI Red Teaming

PointGuard AI performs continuous adversarial testing using Databricks clusters—simulating prompt injections, jailbreaks, and robustness challenges across hundreds of models in parallel. This ensures weaknesses are found and addressed before they can be exploited.

4. AI Detection and Response

With real-time threat monitoring via Structured Streaming, PointGuard AI detects runtime attacks like data leaks, model abuse, and malicious code execution. It can automatically trigger remediation workflows and update threat models using live data streams.

“Security teams need to treat AI as a live environment, not a static deliverable,” said Mali Gorantla. “PointGuard AI delivers full runtime protection, just like XDR does for traditional workloads.”

Pioneering the AISecOps Movement

The DAIS session will also highlight a growing movement toward AISecOps—an end-to-end strategy that treats security as a shared responsibility throughout the AI development lifecycle. Just as DevSecOps revolutionized application security by embedding controls early in the software pipeline, AISecOps does the same for models, datasets, and inference logic.

PointGuard AI is helping lead this charge, alongside Databricks, by enabling development and security teams to work from a common operating picture of risk, configuration, and compliance.

Built on Databricks, For Databricks

PointGuard AI’s credibility doesn’t stop at its integrations. In fact, the platform itself was built using the Databricks Data Intelligence Platform and DASF as foundational tools. This internal use case demonstrates the power of Databricks to support high-throughput security analytics and rapid solution development.

By consolidating security data operations within Databricks, PointGuard achieved:

• 3× acceleration in feature development
• 50% reduction in development time
• 95% fewer false positives, thanks to enhanced correlation and deduplication

The platform architecture includes:

• Raw Security Data Layer using Delta Lake and Auto Loader
• Processed Intelligence Layer using Feature Store and MLflow
• Analytics Layer for scoring risk and driving enforcement dashboards

This self-reinforcing loop—where Databricks powers PointGuard AI, which in turn secures Databricks environments—makes the solution unique and battle-tested.

Join Us at DAIS 2025

Whether you’re an AI engineer, a security leader, or a compliance officer, the Databricks Data+AI Summit 2025 will offer insights into the future of secure AI operations. Be sure to attend the PointGuard AI session for an exclusive look at how the industry's top innovators are driving secure AI innovation at scale.

The future of AI won’t just be smart—it will be secure. Join us in San Francisco to see how.