AI Projects
Building an AI Content Agent That Cut Production Time by 66%
Built Kesha, an AI content agent that automated repeatable content workflows, reduced time spent on social content creation and editing by 66%, and preserved accuracy and brand alignment through structured guardrails.
Year :
2026
Client :
Pitcher
Challenge
As a content team of one, I wanted to find out how much of my recurring workload I could automate without sacrificing the accuracy, brand alignment, and quality I was responsible for.
Every hour spent on repetitive operations meant I wasn't focused on the human-only aspects of content management that only I could do.
Solution
I built Kesha around structured workflows, persistent instructions, trusted source material, and clear guardrails.
Kesha automated two processes. First was social media content production for LinkedIn.
I trained Kesha to find up-to-date brand content, repurpose it into social copy, and build graphics from scratch in Claude Design. This allowed me to shorten social content production from hours to minutes each week.
I also trained Kesha to perform first-round edits of agency blog content. Agencies don't have the internal systems access of an in-person team member, so I spent a lot of time changing the language to fit internal-only brand and product messaging the agency simply did not have.
By sharing the same information with Kesha, I automated first-round checks for accuracy, messaging, and brand alignment. This allowed me to focus on the overall content quality instead of making the same edits each day.
Kesha also had standard safeguards to keep from inventing statistics, quotes, figures, timeframes, or facts, which made content review and publishing smoother.
Result
66% less time spent on social content creation and editing.
Kesha kept recurring content moving while giving me more time for strategy, messaging, campaigns, and other high-priority work.
Takeaway
The project created a repeatable framework for using AI beyond individual tasks: identify where human judgment adds the most value, then build systems around the work that does not require it.
AI did not replace my role in the content process. It gave me more time to do the parts of it that mattered most.
More Projects
AI Projects
Building an AI Content Agent That Cut Production Time by 66%
Built Kesha, an AI content agent that automated repeatable content workflows, reduced time spent on social content creation and editing by 66%, and preserved accuracy and brand alignment through structured guardrails.
Year :
2026
Client :
Pitcher
Challenge
As a content team of one, I wanted to find out how much of my recurring workload I could automate without sacrificing the accuracy, brand alignment, and quality I was responsible for.
Every hour spent on repetitive operations meant I wasn't focused on the human-only aspects of content management that only I could do.
Solution
I built Kesha around structured workflows, persistent instructions, trusted source material, and clear guardrails.
Kesha automated two processes. First was social media content production for LinkedIn.
I trained Kesha to find up-to-date brand content, repurpose it into social copy, and build graphics from scratch in Claude Design. This allowed me to shorten social content production from hours to minutes each week.
I also trained Kesha to perform first-round edits of agency blog content. Agencies don't have the internal systems access of an in-person team member, so I spent a lot of time changing the language to fit internal-only brand and product messaging the agency simply did not have.
By sharing the same information with Kesha, I automated first-round checks for accuracy, messaging, and brand alignment. This allowed me to focus on the overall content quality instead of making the same edits each day.
Kesha also had standard safeguards to keep from inventing statistics, quotes, figures, timeframes, or facts, which made content review and publishing smoother.
Result
66% less time spent on social content creation and editing.
Kesha kept recurring content moving while giving me more time for strategy, messaging, campaigns, and other high-priority work.
Takeaway
The project created a repeatable framework for using AI beyond individual tasks: identify where human judgment adds the most value, then build systems around the work that does not require it.
AI did not replace my role in the content process. It gave me more time to do the parts of it that mattered most.
More Projects
AI Projects
Building an AI Content Agent That Cut Production Time by 66%
Built Kesha, an AI content agent that automated repeatable content workflows, reduced time spent on social content creation and editing by 66%, and preserved accuracy and brand alignment through structured guardrails.
Year :
2026
Client :
Pitcher
Challenge
As a content team of one, I wanted to find out how much of my recurring workload I could automate without sacrificing the accuracy, brand alignment, and quality I was responsible for.
Every hour spent on repetitive operations meant I wasn't focused on the human-only aspects of content management that only I could do.
Solution
I built Kesha around structured workflows, persistent instructions, trusted source material, and clear guardrails.
Kesha automated two processes. First was social media content production for LinkedIn.
I trained Kesha to find up-to-date brand content, repurpose it into social copy, and build graphics from scratch in Claude Design. This allowed me to shorten social content production from hours to minutes each week.
I also trained Kesha to perform first-round edits of agency blog content. Agencies don't have the internal systems access of an in-person team member, so I spent a lot of time changing the language to fit internal-only brand and product messaging the agency simply did not have.
By sharing the same information with Kesha, I automated first-round checks for accuracy, messaging, and brand alignment. This allowed me to focus on the overall content quality instead of making the same edits each day.
Kesha also had standard safeguards to keep from inventing statistics, quotes, figures, timeframes, or facts, which made content review and publishing smoother.
Result
66% less time spent on social content creation and editing.
Kesha kept recurring content moving while giving me more time for strategy, messaging, campaigns, and other high-priority work.
Takeaway
The project created a repeatable framework for using AI beyond individual tasks: identify where human judgment adds the most value, then build systems around the work that does not require it.
AI did not replace my role in the content process. It gave me more time to do the parts of it that mattered most.