Michael
Dunn.
Work
AI workflow integration — embedding AI across design and content production
The challenge
Creative and content teams were under increasing pressure to produce more, faster — without increasing headcount or sacrificing quality. AI tools were emerging rapidly but most organisations were experimenting reactively, without a clear framework for evaluating, embedding, or governing them meaningfully.
Part one: what’s embedded and proven
I identified, evaluated, and embedded AI tooling across the design and content production workflow — focusing on areas where AI could have immediate, measurable impact.
Riverside.FM — embedded into the video and podcast production workflow for thought leadership content. AI-powered recording, editing, transcription, and clip generation significantly reduced post-production time and enabled the team to produce studio-quality content without a production crew.
Bynder AI search — implemented AI-powered asset discovery within the DAM platform, making it significantly faster for teams to find and reuse approved brand assets. Reduced time spent searching and improved the reuse rate of existing content.
Adobe Firefly — introduced for image retouching, background generation, and creative exploration. Enabled the design team to move faster on imagery without relying on photography shoots for every piece.
For each tool: evaluated against real workflow needs, piloted with the team, built usage guidelines and governance, trained the team to integrate it into day-to-day practice.

Part two: what’s in active exploration
AI is moving faster than any organisation can fully absorb. My approach is to stay ahead of it — actively testing emerging tools and building understanding of where they're heading, so that when the right moment comes to commit, the groundwork is already there.
Claude Design (research preview) — currently testing Anthropic's Claude Design for AI-assisted visual design creation. Evaluating its potential for generating on-brand marketing assets — social tiles, infographics, and presentation content — directly from a brief, with brand parameters built in.
Claude AI + Figma MCP — exploring the integration of Claude with Figma via Model Context Protocol, enabling AI-assisted design generation within the design system itself. Early testing focused on the creation of branded marketing assets — social content, infographics, and presentations — produced faster and with more consistent output than a traditional human-only workflow.
My conviction: an AI-supported production team — working with the right tools, the right brand system, and the right governance — will produce more consistent output and work significantly faster than a traditionally structured team. The infrastructure is being built now. The results will follow.
What I've delivered — and what's coming
Delivered:
In progress:
The outcome so far
AI is now embedded across multiple stages of the design and content production workflow. The team moves faster, produces more, and maintains brand quality. The exploration work underway is laying the groundwork for a step-change in what a lean creative team can produce.
Results
Discipline
AI and tools
My role
Evaluation
Piloting
Implementation
Exploration
Practices
RiversideFM
Bynder AI
Adobe Firefly
Claude
Figma MCP
Skills
AI tool evaluation
Workflow integration
Governance and guidelines
Team enablement
Impact
Faster production across video, imagery and assets
Future-state AI program
Brand identity
Rebranding newly combined firm Norton Rose Fulbright#RiskReady — McLaren Racing partnership activationGlobal brand advertising — top 3 global legal brandDesign systems
Office design system — PowerPoint & Word at global scaleModernising website visual design and experience© Michael Dunn — 2026
Work
AI workflow integration — embedding AI across design and content production
The challenge
Creative and content teams were under increasing pressure to produce more, faster — without increasing headcount or sacrificing quality. AI tools were emerging rapidly but most organisations were experimenting reactively, without a clear framework for evaluating, embedding, or governing them meaningfully.
Part one: what’s embedded and proven
I identified, evaluated, and embedded AI tooling across the design and content production workflow — focusing on areas where AI could have immediate, measurable impact.
Riverside.FM — embedded into the video and podcast production workflow for thought leadership content. AI-powered recording, editing, transcription, and clip generation significantly reduced post-production time and enabled the team to produce studio-quality content without a production crew.
Bynder AI search — implemented AI-powered asset discovery within the DAM platform, making it significantly faster for teams to find and reuse approved brand assets. Reduced time spent searching and improved the reuse rate of existing content.
Adobe Firefly — introduced for image retouching, background generation, and creative exploration. Enabled the design team to move faster on imagery without relying on photography shoots for every piece.
For each tool: evaluated against real workflow needs, piloted with the team, built usage guidelines and governance, trained the team to integrate it into day-to-day practice.

Part two: what’s in active exploration
AI is moving faster than any organisation can fully absorb. My approach is to stay ahead of it — actively testing emerging tools and building understanding of where they're heading, so that when the right moment comes to commit, the groundwork is already there.
Claude Design (research preview) — currently testing Anthropic's Claude Design for AI-assisted visual design creation. Evaluating its potential for generating on-brand marketing assets — social tiles, infographics, and presentation content — directly from a brief, with brand parameters built in.
Claude AI + Figma MCP — exploring the integration of Claude with Figma via Model Context Protocol, enabling AI-assisted design generation within the design system itself. Early testing focused on the creation of branded marketing assets — social content, infographics, and presentations — produced faster and with more consistent output than a traditional human-only workflow.
My conviction: an AI-supported production team — working with the right tools, the right brand system, and the right governance — will produce more consistent output and work significantly faster than a traditionally structured team. The infrastructure is being built now. The results will follow.
What I've delivered — and what's coming
Delivered:
In progress:
The outcome so far
AI is now embedded across multiple stages of the design and content production workflow. The team moves faster, produces more, and maintains brand quality. The exploration work underway is laying the groundwork for a step-change in what a lean creative team can produce.
Results
Discipline
AI and tools
My role
Evaluation
Piloting
Implementation
Exploration
Practices
RiversideFM
Bynder AI
Adobe Firefly
Claude
Figma MCP
Skills
AI tool evaluation
Workflow integration
Governance and guidelines
Team enablement
Impact
Faster production across video, imagery and assets
Future-state AI program
Brand identity
Rebranding newly combined firm Norton Rose Fulbright#RiskReady — McLaren Racing partnership activationGlobal brand advertising — top 3 global legal brandDesign systems
Office design system — PowerPoint & Word at global scaleModernising website visual design and experience© Michael Dunn — 2026
Work
AI workflow integration — embedding AI across design and content production
The challenge
Creative and content teams were under increasing pressure to produce more, faster — without increasing headcount or sacrificing quality. AI tools were emerging rapidly but most organisations were experimenting reactively, without a clear framework for evaluating, embedding, or governing them meaningfully.
Part one: what’s embedded and proven
I identified, evaluated, and embedded AI tooling across the design and content production workflow — focusing on areas where AI could have immediate, measurable impact.
Riverside.FM — embedded into the video and podcast production workflow for thought leadership content. AI-powered recording, editing, transcription, and clip generation significantly reduced post-production time and enabled the team to produce studio-quality content without a production crew.
Bynder AI search — implemented AI-powered asset discovery within the DAM platform, making it significantly faster for teams to find and reuse approved brand assets. Reduced time spent searching and improved the reuse rate of existing content.
Adobe Firefly — introduced for image retouching, background generation, and creative exploration. Enabled the design team to move faster on imagery without relying on photography shoots for every piece.
For each tool: evaluated against real workflow needs, piloted with the team, built usage guidelines and governance, trained the team to integrate it into day-to-day practice.

Part two: what’s in active exploration
AI is moving faster than any organisation can fully absorb. My approach is to stay ahead of it — actively testing emerging tools and building understanding of where they're heading, so that when the right moment comes to commit, the groundwork is already there.
Claude Design (research preview) — currently testing Anthropic's Claude Design for AI-assisted visual design creation. Evaluating its potential for generating on-brand marketing assets — social tiles, infographics, and presentation content — directly from a brief, with brand parameters built in.
Claude AI + Figma MCP — exploring the integration of Claude with Figma via Model Context Protocol, enabling AI-assisted design generation within the design system itself. Early testing focused on the creation of branded marketing assets — social content, infographics, and presentations — produced faster and with more consistent output than a traditional human-only workflow.
My conviction: an AI-supported production team — working with the right tools, the right brand system, and the right governance — will produce more consistent output and work significantly faster than a traditionally structured team. The infrastructure is being built now. The results will follow.
What I've delivered — and what's coming
Delivered:
In progress:
The outcome so far
AI is now embedded across multiple stages of the design and content production workflow. The team moves faster, produces more, and maintains brand quality. The exploration work underway is laying the groundwork for a step-change in what a lean creative team can produce.
Results
Discipline
AI and tools
My role
Evaluation
Piloting
Implementation
Exploration
Practices
RiversideFM
Bynder AI
Adobe Firefly
Claude
Figma MCP
Skills
AI tool evaluation
Workflow integration
Governance and guidelines
Team enablement
Impact
Faster production across video, imagery and assets
Future-state AI program
Brand identity
Rebranding newly combined firm Norton Rose Fulbright#RiskReady — McLaren Racing partnership activationGlobal brand advertising — top 3 global legal brandDesign systems
Office design system — PowerPoint & Word at global scaleModernising website visual design and experience© Michael Dunn — 2026