Independent product
Kumosei: AI for content and marketing
Kumosei applies AI to content and marketing workflows. As co-founder and CTO, I lead the product’s technical direction and the way models and integrations become a usable experience.
The problem
Generating text is only one part of the work. A useful product must account for context, consistency, review and the hand-offs between the activities that make up the workflow.
The technical challenge is to integrate probabilistic capabilities into clear journeys while keeping inputs, responsibilities and control explicit.
The solution and my contribution
- Technical direction and architectural choices.
- Integration of AI capabilities into the product.
- Translation of use cases into flows and requirements.
- Platform evolution alongside business priorities.
The approach
AI is treated as one component of the system, not the whole product. Interface, data, controls and integrations determine whether the result is genuinely useful.
Decisions favour short feedback cycles and observable journeys, distinguishing a convincing demo from a capability that works in everyday operations.
The concrete result
A SaaS platform for content creation, editorial management and digital publishing, with multilingual and multi-site workflows. The project is evolving; the public website introduces the product and offers access on request.
Visit Kumosei (opens in a new tab)Technology with a purpose
- Next.js and Node.js for the application and services.
- PostgreSQL and Prisma for data and database access.
- LLM integrations and SEO automation for editorial workflows.
- Docker for the application environment.