AI basics
To confidently teach Dragon Copilot, you need to understand the fundamental AI concepts behind the technology. A clear grasp of basic AI terminology helps you explain how Dragon Copilot works and improves your ability to support clinicians as they adopt AI-assisted workflows.
Note
This unit applies to Dragon Copilot and DAX Copilot. The features and functionality available to you vary based on your organization's configuration, language settings, and whether you're using the web, mobile, desktop, or embedded app.
AI terms
Dragon Copilot uses artificial intelligence to capture, interpret, and generate clinical content. Knowing these terms strengthens your ability to support clinicians' understanding.
Here are some foundational AI terms used in Dragon Copilot:
Artificial intelligence (AI): Artificial intelligence refers to computer systems designed to perform tasks that typically require human intelligence, such as understanding speech, make decisions, translating language, or recognize patterns. AI systems learn from large amounts of data and follow algorithms to complete tasks. AI can work behind the scenes (like predictive text) or through direct interaction (like a chatbot or Copilot experience).
Machine learning: Machine learning is a subset of AI that uses large datasets to train algorithms and create models capable of tasks such as image recognition or language translation. These systems learn patterns from data and improve over time.
Large language models (LLMs): Large language models use machine learning to learn language patterns from vast amounts of text. They can answer questions, summarize information, translate languages, and produce human-like responses. Developers often fine-tune LLMs for specific use cases.
Conversational AI: Conversational AI enables computers to understand, interpret, and respond to human speech. Virtual assistants and chatbots are examples. Dragon Copilot uses conversational AI trained on billions of clinical narrative minutes to recognize clinical speech and generate clinically relevant output.
Generative AI: Generative AI can create new content, such as text, images, or summaries, by learning from patterns in existing data. Dragon Copilot uses finely tuned generative AI to efficiently create accurate clinical notes, referral letters, after visit summaries, and other documentation.
Ambient AI: Ambient AI works unobtrusively in the background. In Dragon Copilot, ambient AI securely captures clinician–patient conversations in real time, so clinicians can focus on care while AI transforms the conversation into high-quality documentation.
Responsible AI: Responsible AI ensures technologies are designed and used safely, ethically, and equitably. It covers system design, data handling, user experience, and access controls. Microsoft’s AI solutions, including Dragon Copilot, adhere to responsible AI principles to ensure safety, privacy, and trust.
Copilot: A copilot is an AI-powered digital assistant that helps with tasks such as writing, summarizing, searching, and analyzing information. It responds to natural language and is built with responsible AI guardrails. Like a copilot in aviation, it assists the user, but the user remains in control.
Prompts: A prompt is an instruction (spoken, typed, or structured) that tells the AI what task to perform. In Dragon Copilot, verbalization best practices, specialty-specific outputs, and user input shape prompts to ensure content is generated accurately.
How AI works in Dragon Copilot
Dragon Copilot combines the dictation capabilities of front-end speech applications (like Dragon Medical One) and ambient listening capabilities (introduced in DAX Copilot) into a single solution. The underlying technology uses responsible AI principles to bring together machine learning, large language models, conversational AI, ambient AI, and finely tuned generative AI to help clinicians manage their clinical workflow.
When you understand how Dragon Copilot uses AI, you can explain the technology and answer questions in simple, clinician-friendly terms.
High-level Dragon Copilot workflow
- Dragon Copilot captures the clinician-patient conversation through ambient recording.
- The AI processes the audio by using finely tuned models.
- The AI generates structured documentation.
Several factors influence Dragon Copilot's output, including:
- Verbalization. What the clinician says and how they say it impacts the AI output. When a clinician follows verbalization best practices, they achieve better results and spend less time editing.
- Specialty and role settings. Each primary specialty has settings that apply when the AI model analyzes the input. This configuration includes verbiage, sections, and other formatting. Set your primary specialty when you open Dragon Copilot for the first time. You can always change this setting later.
- Style and format preferences. Set style and formatting preferences in Settings
. Select bulleted list versus paragraphs or a customizable template for each note section. Dragon Copilot uses style and format preferences as a guide to build the AI-generated summary.
AI features
Clinicians can use the chat box to ask questions, summarize a note, make edits, add context, or generate other documentation like patient instructions or referral letters. They provide this information through prompts.
In simple terms, Dragon Copilot acts like a clinical assistant, listening, interpreting, and generating documentation to help clinicians work more efficiently.