Hey everyone! :wave: We from Flock have been work...

# feed

[Willem Veelenturf]

Hey everyone! 👋

We from Flock have been working on a Kotlin multiplatform library called Aigentic for building AI agents, and we would love to get some feedback from the community.

It allows you to create AI agents using a clean, native DSL. If you're looking to integrate LLMs into your Kotlin apps, we think you'll find it useful.

We are in the early stages, so any thoughts on the API, documentation, or overall experience would be incredibly valuable.

- 📚 Check out the docs here: https://aigentic.io/
- ⭐ GitHub: https://github.com/flock-community/aigentic
- 🏃‍➡️ Init: https://github.com/flock-community/aigentic-initializr

[joseph_ivie]

Given the existence of the Jetbrains-official Koog, it would help if your README explained the motivation, comparative benefits, and perhaps even a direct example comparison.
Don't let me discourage you guys though; my project (KiteUI) is competing with Compose for heaven's sake. Just help me understand it relative to the standard.

[nsmnds]

Thanks for the great question!

Aigentic is focused just on business process automation, whereas Koog seems like a broader, more all encompassing framework that tackles many different types of agent applications.

In our experience, the technical setup (DSL, framework) is only about 20% of building production ready agents. The other 80% is the data science work - evaluation, performance monitoring, bias detection, edge case handling, etc. Questions like: How do I evaluate performance and stability? Is my test set biased? How many test cases do I need for confidence? How do I improve configs for edge cases?

The DSL is just our first step. We're also developing a platform (planning to open source) that tackles that challenging 80%. We're making it agent DSL agnostic, so it'll probably work with Koog and other frameworks too, not just our own DSL.