Peter Ward

Extending Microsoft Copilot: Visual Intelligence with Midjourney

Written by Peter Ward | May 17, 2026 3:49:53 PM

 

 
 

 

Earlier this month, I presented at M365 Philly on integrating Copilot and the image AI tool Midjourney.

Midjourney is not a Microsoft product, but is fantastic at producing images based on prompts.

 

Other Key features

  • Text-to-image generation – Converts written prompts into artwork
  • Highly artistic style – Known for cinematic, stylized, and aesthetically strong images
  • Variations & upscaling – You can refine or enhance generated images
  • Prompt-based control – You guide the output through wording, style tags, and parameters

 

 

 How it works

Midjourney uses advanced diffusion models (similar to DALL·E or Stable Diffusion) trained on vast datasets of images and text to “imagine” new visuals based on your input.

 

How people use it

  • Marketing visuals and campaigns
  • Concept art (games, movies, products)
  • Social media content
  • Mood boards / creative exploration
  • Branding and design inspiration

 

How does it work with Microsoft Copilot?

Firstly, there's no native integration with connectors

  • Midjourney → independent AI image tool (runs via Discord + web app)
  • Microsoft Copilot → productivity AI embedded in Microsoft 365, Edge, Windows, etc.Microsoft uses its own image model (DALL·E / OpenAI) inside Copilot.. .

 

 Where Copilot is very good, it is building prompts for Midjourney.

The exercise I presented on how Copilot and Midjourney work together is that I asked someone in the audience to mention a movie... and they said - Gone in 60 Seconds. Then put someone in the audience to give me their Linkedin profile and we'll put their face on the poster.

 

Here are the steps:

In Copilot: Describe this movie poster?

  

 

>> 

The poster is dominated by an extreme close-up portrait of a person’s face, filling most of the frame. The composition is tightly cropped, with the face slightly off-center, drawing strong attention to the eyes. The lower-right portion of the image introduces a secondary visual element—a blurred, fast-moving car.

 

Foreground elements

1. The face (primary focal point)

  • The image shows a highly detailed close-up of a human face, occupying roughly 70–80% of the poster.
  • The eyes are sharply in focus and positioned near the upper center, making them the main anchor point.
  • The facial features are lit with strong, warm lighting, creating pronounced highlights and shadows.
  • The overall texture appears grainy and stylized, giving a gritty, cinematic feel.

2. The car (secondary focal point)

  • In the lower-right corner, there is a sports car rendered with motion blur, suggesting speed.
  • The car is angled diagonally, moving from right to left across the frame.
  • Bright reflections and streaks of light emphasize motion and energy.

🎨 Colors and lighting

  • The poster uses a warm, high-saturation palette:
    • Dominant tones: orange, gold, and amber
    • Accents: red and white in the text
  • Lighting is intense and directional:
    • The face is illuminated from one side, creating contrast and depth
    • The car reflects bright highlights, adding a metallic sheen
  • The overall effect is fiery and high-energy, reinforcing action and intensity.

 

  

>>If this was a midjounrey prompt what would it be?

 

 extreme close-up portrait of a person’s face dominating the frame, sharp detailed eyes in focus, gritty skin texture, warm orange and gold color grading, dramatic side lighting, intense cinematic look, motion-blurred sports car streaking across the lower corner, strong sense of speed and urgency, high contrast highlights, glowing reflections, action movie poster style, bold composition, shallow depth of field, 85mm lens, film grain --ar 2:3 --v 6 --style raw 

 

Put the prompt in Midjourney. This was the output

 

 

After prompting in Midjourney with the face in the audience, this was the result

 

 

 
View full post