I posted a couple of months ago on building a low code AI ML platform, and since then I've baked in a lot of new stuff. First of all, what is otto-m8? The best way to think about it is its a low code platform that let's you visually create AI/ML based workflows by treating building AI/ML flows as a graph problem. Every tech meetup I attend, I've been seeing everyone demo-ing their workflow via what looks like a graph or flowchart. Hence why I decided to give it a shot and make those flowcharts do something.

What has changed since my last post? I added the ability for devs to alter an implementation of a block, which essentially helps you do two things:

- create custom blocks from scratch, or

- not be limited to what I have implemented.

Furthermore, I added features for LLM memory, saving drafts, testing workflows on the fly without having to deploy it(kind of mimics a real life dev workflow with instant feedbacks on your changes), a GCloud integration and many more.

Yes, I'm aware of Langflow and some other startups working on the same problem space(and how could one forget n8n, from whom this project was inspired). The way I like to think of otto-m8 is that it is not built on top of any frameworks like Langchain which most of these low code platforms are built on top of. This gives developers the flexibility to not implement or be tied to Langchain if they don't want to while conversely allowing them to use Langchain if they want to. I don't know if that's a good thing but I like to personally just implement the cookie cutter OpenAI API starter code, and iterate.

It's not fully worked on yet, and hence why its open source- there's only so much I could do alone. So I'm basically hoping that I could grow a community around it, and make something cool. So I hope you guys like it!

Here's some tutorials: https://otto-m8.com/docs/category/tutorials