Building now
An AI career product built around decisions, not document generation.
I built Grapevines solo, end to end. It combines adaptive conversation, research, voice calibration, and structured evaluation so a professional can reason through a career move before generating materials.
- Role
- Founder and solo builder
- System
- Conversation, research, evaluation, and orchestration
The working product joins adaptive conversation, research, voice calibration, and a scoring pipeline in one operating system.
The decision comes before the document.
Most career software starts with an output. Grapevines starts with the decision underneath it: what the person is trying to change, what evidence supports that move, and who will evaluate it.
Resumes, outreach, and interview material are downstream artifacts. The product first establishes positioning, researches the context, and makes the reasoning inspectable.
The system has to know when it does not know.
Research and evaluation use explicit fallbacks. If the system cannot establish a person, company detail, or fit judgment, it returns the gap instead of inventing an answer.
That boundary matters more than making every screen look complete. The product is designed to keep a human responsible for consequential career decisions.
What I am testing now.
The next question is not whether more material can be generated. It is whether the system helps someone make a better decision and act on it with less wasted effort.
Interface views
The operating boundary, on screen.
