Mukund Asagodu | Design Leadership

Case Study

Author Assist

Elevating content quality at scale with AI.

Role

Design Director

Product

Internal CMS Solution

Teams

Design, Product, Engineering, Content, Research. Testing, Data

Year

2025

Challenge

Improve speed and consistency at scale without increasing risk or making authoring feel opaque.

Solution

Clear, actionable AI suggestions that improve content quality by aligning with the writing style guide.

Impact

150+ authors supported, higher content quality at scale, faster authoring workflows, fewer style guide violations, reduced review friction, safety signals for high-impact edits, and reduced customer support load.

A closer look at the problem.

What we saw was consistent. The challenge was not just speed, but introducing AI without increasing risk or making authoring feel opaque.

Authors spent too much time on repetitive editing and formatting, especially during high-volume updates.

Reviewers struggled to spot meaningful changes and enforce consistency across thousands of articles.

Inconsistency was not just a quality issue. It affected customer comprehension and increased support load.

Authors spent too much time on repetitive editing and formatting, especially during high-volume updates.

Reviewers struggled to spot meaningful changes and enforce consistency across thousands of articles.

Inconsistency was not just a quality issue. It affected customer comprehension and increased support load.

Authors spent too much time on repetitive editing and formatting, especially during high-volume updates.

Reviewers struggled to spot meaningful changes and enforce consistency across thousands of articles.

Inconsistency was not just a quality issue. It affected customer comprehension and increased support load.

What we designed.

We designed Author Assist to support authors without disrupting their flow. The goal was to keep authors in control, not hand control to AI.

Suggestions that are prioritized, organized and easy to scan.

Human decisions stay central, with clear actions.

Flagged, explained, and mapped to the styleguide.

Safety signals designed for trust. Built for confidence.

Inline track changes with visual clarity.

Suggestions that are prioritized, organized and easy to scan.

Human decisions stay central, with clear actions.

Flagged, explained, and mapped to the styleguide.

Safety signals designed for trust. Built for confidence.

Inline track changes with visual clarity.

Suggestions that are prioritized, organized and easy to scan.

Human decisions stay central, with clear actions.

Flagged, explained, and mapped to the styleguide.

Safety signals designed for trust. Built for confidence.

Inline track changes with visual clarity.

Impact.

Content, authored in the tool, powers customer and advisor support experiences at global scale. Author Assist became an early proof point that AI can boost speed and quality while preserving trust.

Scaled author support.

150+ authors supported across 4,000+ articles and 33 language sets with consistent in-editor guidance.

Faster draft-to-publish.

Reduced cycle time by cutting manual formatting and rework, and making suggestions easier to review, validate, and apply.

Higher quality at scale.

Improved style guide adherence across content, resulting in fewer violations per article and more consistent writing standards.

Improved self-solve success.

Delivered clearer, more consistent support content that customers could understand quickly and act on with confidence.

Lower downstream support burden.

Improved clarity and consistency contributed to lower assisted-support volume and operational cost.

Scaled author support.

150+ authors supported across 4,000+ articles and 33 language sets with consistent in-editor guidance.

Faster draft-to-publish.

Reduced cycle time by cutting manual formatting and rework, and making suggestions easier to review, validate, and apply.

Higher quality at scale.

Improved style guide adherence across content, resulting in fewer violations per article and more consistent writing standards.

Improved self-solve success.

Delivered clearer, more consistent support content that customers could understand quickly and act on with confidence.

Lower downstream support burden.

Improved clarity and consistency contributed to lower assisted-support volume and operational cost.

Scaled author support.

150+ authors supported across 4,000+ articles and 33 language sets with consistent in-editor guidance.

Faster draft-to-publish.

Reduced cycle time by cutting manual formatting and rework, and making suggestions easier to review, validate, and apply.

Higher quality at scale.

Improved style guide adherence across content, resulting in fewer violations per article and more consistent writing standards.

Improved self-solve success.

Delivered clearer, more consistent support content that customers could understand quickly and act on with confidence.

Lower downstream support burden.

Improved clarity and consistency contributed to lower assisted-support volume and operational cost.

Behind the scenes.

A read of my contributions and the cross-functional collaboration that took Author Assist from a focused ask to a shipped first phase and a roadmap teams could build from next.

Read the full story