Case Study: AI Systems That
Give the Team Time Back
At a glance
- Client
- Sage Growth Partners
- Industry
- Healthcare marketing agency
- My role
- Manager, Digital Marketing & Analytics · Sage Growth Partners
- Timeline
- 2025 – present
- Services
- AI & automation, Paid media, Analytics
Headline results
- Cut client service hours on reporting and campaign work 35% with AI systems built end to end
- Trained 12 team members on Claude and other LLMs in hands-on workshops
- Built a custom connector that lets Claude make changes in Google, Microsoft and LinkedIn Ads
Summary
At Sage Growth Partners, a healthcare marketing agency, I built the agency’s AI systems end to end and got the team using them. The work cut client service hours on reporting and campaign work 35%.
I built the agent workflows and tool integrations myself: the skills, the subagents and the connectors that tie them to the tools the team already uses. Because these systems were built for Sage and its clients, the details here stay general.
What I built
- A team-wide client database. One shared home for client information, used across the team.
- Automated performance reporting. Performance reports that build themselves instead of being pulled together by hand.
- Agentic campaign management. AI agents that take on campaign management across Google Ads, LinkedIn Ads and Microsoft Advertising (Bing).
- Brand-locked design systems. Systems that generate on-brand decks and sales collateral.
- A custom ads connector. An MCP connector I built so I can make changes in Google Ads, Microsoft Advertising and LinkedIn Ads directly through Claude.
Getting the team to use it
Tools only save time if people use them. I designed and ran hands-on workshops that trained 12 team members on Claude and other LLMs. I also advise clients on bringing AI into their own marketing.
AI in the rest of my work
- Search that AI answers. For a neurosurgical practice, appearances in AI-generated search results rose from 13 in Q1 2025 to 2,250 in Q1 2026, and keywords showing up in AI results rose 5,953%. Q1 2025 is the baseline from before I started. The full story is in the New Jersey Brain & Spine case study.
- Building with AI. Stampede is an AI-directed build: I took the app from idea to the App Store, with Claude Code handling implementation while I directed the product, reviewed the architecture and ran the deploys.
- Training. I completed Northwestern Kellogg’s Digital Marketing Strategies: Data, Automation, AI & Analytics program in 2025.
The results in numbers
The program
- Fewer service hours on reporting and campaign work
- 35%
- Team members trained
- 12
- Ad platforms in the custom connector
- 3