Superesque · TypingMind

The problem

Non-technical professionals need to learn how to build AI agents and design AI systems. Superesque needed a platform to facilitate this education and run its operations with AI agents.

What was deployed

Vendor

TypingMind

Products

TypingMind, TypingMind Teams

Models

Gemini

Technique

LLM

Build or buy

Bought off the shelf

Deployment

SaaS

Data used

external RAG knowledge base; Notion for external RAG

Scale

piloted three cohorts this year

What changed

Each row is quoted from the source. Figures we could not find on the page in those words are marked — they are kept, not deleted, so you can judge them.

Difficulties and limits

• I find it frustrating how I can't carefully select what contents I sync and also how difficult it is to edit the synced content in TypingMind. Once you've pulled a volume of stuff into the knowledge base, things get quite messy.
• I'd like to see improvements to the knowledge base. A RAG system is such an important part of AI building. I don't need an on-platform vector database like Cassidy. But I would appreciate the ability to structure the contents of my knowledge base a bit more and organize things into folders and subfolders.
• Also, for Typingmind teams, it would be great if we could give students a personal knowledge base. Even if it just had minimal storage data, it would mean I could train them on how to build and use a knowledge base, which is super important from an educational point of view.

• Superesque is still in the pilot stage so we don't have any mind-blowing metrics to share.

Sources

• typingmind.com — vendor self-report