#Where to start ?

6 messages · Page 1 of 1 (latest)

crude tapir
#

Hello all,

I'm a seasoned developer and I would like to start "doing some AI related things", mostly integrated in services or for automation, so likely agent coding and orchestration from my understanding.

I'm kind of lost between all the acronyms and tools like langchain/graph, haystack, n8n, semantic kernel and all.

I woul ideally like to start from scratch to understand the concepts properly.
My idea right now is to start with plain C# and ollama to interact with an LLM and slowly learn the standards around it.

If anyone has a good link to start (ideally from scratch) and understand it all, I'm very interested, thanks 🙂

sick swift
# crude tapir Hello all, I'm a seasoned developer and I would like to start "doing some AI re...

I really like your approach brother — starting from scratch and wanting to actually understand how things work is the smartest way to go. I went down a similar path: started with Google’s ML Crash Course, then gradually moved into LLMs and orchestration tools like LangChain (still exploring!).

One heads-up though — while using Ollama is a great call, doing everything in C# might slow you down a bit. Most of the LLM tooling ecosystem (LangChain, transformers, agent frameworks, vector DB integrations) is very Python-centric. You can absolutely build in C#, especially with things like Semantic Kernel, but it might make the learning curve steeper just due to lack of examples and community resources.

Personally, I’m learning and teaching this stuff on YouTube as I go — and Python has made it way easier to experiment and test ideas quickly.

That said, your idea to build from the ground up and gradually layer in orchestration is spot on. If you ever want to bounce around ideas or resources, happy to chat!

crude tapir
sick swift
#

All best brother 👍

steady gulch
#

im already done with CNN