Fig. 6 Research app
Personal AI outreach with full cost visibility
A GHL cleanup plus a custom AI app that researches each company and writes 4 tailored emails per company, with AI costs logged on every run.
- Client
- B2B analytics firm
- Industry
- B2B analytics
- Year
- 2026
- Discipline
- AI workflows, Custom apps, GHL systems
Drawing
How it works
Step 01
Companies pulled from GHL
Step 02
Researched with Gemini, Tavily and SERP
Step 03
4 emails written per company
Step 04
Saved back to GHL and sent
The problem
The firm left HubSpot after AI enrichment charges kept climbing with no clear reason why. They needed a clean, structured CRM, and outreach that felt personal without hiding what it cost.
Built with
- GoHighLevel
- React
- TypeScript
- Gemini API
- Tavily
- SERP API
What I built
- 01A full GHL audit and data cleanup, plus contact and company scoring to rank the pipeline.
- 02Sales pipelines structured from discovery through closing and delivery.
- 03A custom outreach app that pulls companies and contacts from GHL, researches each company with Gemini, Tavily and SERP, and writes 4 emails per company tailored to each contact's role.
- 04Research and emails saved back to GHL and sent from it, so the CRM stays the single source of truth.
- 05Token usage logged on every research and writing run, so AI spend is visible per operation.
Next drawing
A bilingual buyer pipeline with AI property matching