Microsoft Certification Propels BastionGPT into Healthcare's Automation Core

DALLAS, TX – September 29, 2026 — The race to automate administrative workflows in healthcare has long been hampered by a critical bottleneck: the strict data privacy requirements of the Health Insurance Portability and Accountability Act (HIPAA). While low-code platforms have revolutionized how enterprise IT departments build internal tools, integrating advanced generative artificial intelligence into these systems has remained a compliance minefield. Today, Bastion Intelligence announced a significant breakthrough, launching a Microsoft-certified Power Platform connector that brings its HIPAA-compliant AI directly into Power Automate, Power Apps, Copilot Studio, and Azure Logic Apps.

This integration is not merely a feature update for the software provider; it represents a fundamental unblocking of enterprise automation in the clinical space. By securing official listing in Microsoft’s connector directory and documentation on Microsoft Learn, the Dallas-based firm is solving one of the most persistent friction points for healthcare IT administrators: deploying multi-model large language models (LLMs) without compromising patient data security or taking on the burden of custom API maintenance.

Closing the Compliance Loophole in Low-Code Automation

In recent years, low-code tools like Microsoft Power Automate have proliferated across hospital back offices, enabling staff to build automated workflows for everything from email triage to document routing. However, when these organizations attempt to inject generative AI into these flows—for instance, to summarize a patient's medical history or extract key data from a referral PDF—they hit a regulatory wall.

The issue lies in the Business Associate Agreement (BAA), a mandatory contract under HIPAA that dictates how protected health information (PHI) is handled by third-party vendors. While Microsoft’s standard BAA covers its own in-scope services, including the Power Platform infrastructure, it does not automatically extend to external AI services connected via third-party APIs. If a hospital uses a basic HTTP connector to send patient data to a public LLM, they are likely in breach of HIPAA.

The newly certified connector bridges this exact legal and regulatory gap. Every action executed through the connector calls the application programming interface under a comprehensive BAA with the AI provider.

"In healthcare automation, every service that processes patient information needs a BAA," said Joshua Spencer, CEO of Bastion Intelligence, in the company's press release. "Microsoft’s covers Power Automate, and ours covers the AI. Many of our customers already run their practices on Microsoft 365 and have asked how to put BastionGPT to work inside the flows and agents they use every day. Now they can, without building and maintaining a custom API integration."

Crucially, the agreement contractually guarantees that customer content—including prompts, uploaded documents, and audio recordings—is never used to train the underlying AI models, shielding healthcare organizations from the data leakage controversies that have plagued consumer-grade AI tools.

Multi-Model LLM Orchestration in the Clinical Back Office

Beyond compliance, the technical capability of the new connector introduces a sophisticated approach to AI deployment known as multi-model orchestration. Rather than locking users into a single AI ecosystem, the platform dynamically routes requests to the best-fit model among the latest iterations of OpenAI's GPT, Anthropic's Claude, and Google's Gemini.

This agnostic approach is highly strategic. Different LLMs excel at different tasks; one might be superior at complex clinical reasoning, while another is faster and more cost-effective for basic text summarization. By handling this routing automatically beneath the surface—all while maintaining the same BAA umbrella—the platform ensures optimal performance without requiring IT staff to manage multiple vendor relationships. Users who have a specific preference can also manually select their desired model.

The connector introduces five core actions through Microsoft’s visual workflow tools. Users can ask questions with optional document references, create multi-turn chat completions with custom system messages, upload reference documents up to 10 megabytes for 30-day retention, submit audio files for transcription with speaker labels, and retrieve completed transcripts as both structured segments and full text.

In practice, these actions enable highly complex automated clinical workflows. For example, a hospital can configure a Power Automate flow that detects when a new referral packet arrives in a SharePoint folder. The flow can automatically upload the PDF, ask the AI to extract the referring provider’s details and the requested medical service, and route a summarized alert to the triage team for review. Another flow could monitor a OneDrive folder for recorded telemedicine sessions, automatically submit the audio for transcription, and save the timestamped, speaker-labeled text alongside the original file.

Navigating a Crowded Healthcare AI Ecosystem

The launch of this certified connector occurs within a fiercely competitive healthcare AI landscape. Tech giants and specialized startups alike are vying for dominance in the clinical documentation space. Microsoft itself offers a robust suite of tools, including the Microsoft Cloud for Healthcare, Azure OpenAI services, and Nuance DAX Copilot.

However, the Dallas-based company has carved out a distinct market position by focusing on broad accessibility and avoiding deep, cumbersome integrations with Electronic Health Record (EHR) systems. While products like Nuance DAX Copilot offer powerful, point-of-care automation deeply embedded within specific EHRs, they often come with steep price tags and complex deployment cycles.

In contrast, this platform operates agnostically. It can be used alongside any major EHR—from Epic and Cerner to specialized platforms like TherapyNotes—via simple copy-paste mechanisms, document uploads, or now, through Microsoft 365 automations. This flexibility has driven significant market adoption. The company reports that its platform is trusted by more than 10,000 healthcare organizations, a claim supported by its high ratings across major B2B software directories.

The clinical efficacy of the tool has also been independently validated. A recent peer-reviewed study conducted by physicians at Baylor College of Medicine and Texas Children's Hospital evaluated the platform's performance in a pediatric intensive care unit, finding a 99.3 percent sentence-level accuracy rate in its clinical documentation. This level of proven reliability is critical for IT leaders tasked with justifying the integration of generative AI into high-stakes clinical environments.

The Economics of Specialized Connectors

The introduction of this tool also highlights a broader trend in the enterprise software ecosystem: the rise of domain-specific, certified connectors in the Power Platform marketplace. By achieving official certification, independent software vendors can embed their services directly into the daily workflows of millions of Microsoft users.

The economics of this specific integration are straightforward but require careful planning from IT procurement teams. The connector is classified as a Premium connector for Power Automate, Power Apps, and Copilot Studio, and as a Standard connector for Azure Logic Apps. It is available in supported commercial cloud regions worldwide, though notably excluded from government and China cloud environments.

To utilize the integration, healthcare organizations must maintain both an active API subscription with the AI provider and the applicable Microsoft licensing that permits Premium connector usage. Usage is billed at the same rates as direct API requests.

While this represents an additional layer of licensing, the return on investment for healthcare organizations is often realized in the dramatic reduction of development and maintenance costs. Building a custom, HIPAA-compliant API integration between an enterprise Microsoft environment and a multi-model LLM architecture requires specialized engineering talent, rigorous security auditing, and ongoing maintenance to accommodate rapidly changing AI endpoints. By packaging this complex infrastructure into a visual, drag-and-drop connector, the company is democratizing access to advanced AI for healthcare operations. Clinical informatics directors and compliance officers can now collaborate to build sophisticated, AI-driven workflows in a matter of hours, rather than months, fundamentally accelerating the pace of digital transformation in the medical sector.