When I first started learning about AI, I mostly thought about it in terms of machine learning models, datasets, training, and accuracy.
But recently, while exploring AWS and preparing myself for the AI Practitioner path, I started looking at AI from a different perspective.
I'm getting more interested in Generative AI and AI agents and how they can actually be used to build applications.
What caught my attention?
One thing I found interesting is Amazon Bedrock.
Instead of building and training a large language model from scratch, Bedrock allows us to work with foundation models through APIs and build applications around them.
This made me realize that building an AI application isn't always about creating a model yourself.
A lot of the challenge is actually about:
- Giving it the right context
- Connecting it with external data
- Managing the application properly
- Evaluating whether the AI is actually giving useful results
AI agents are even more interesting
While exploring GenAI, I came across AI agents.
The idea of an AI system that can understand a goal, decide what steps are needed, use tools, and complete a task is quite different from simply asking a chatbot a question.
For example, instead of:
User → Question → AI → Answer
an agent can work more like:
Goal → Reason → Use tools → Get information → Take action → Result
This is one area I really want to explore more.
What I'm learning right now
I'm still at the learning stage, but I'm trying to connect the concepts I've learned in AI/ML with the AWS ecosystem.
Some areas I'm currently exploring are:
- Building AI applications on AWS
The more I explore, the more I realize that AI is not just about the model anymore. It's about building a complete system around it.
I'm looking forward to learning more, experimenting with small projects, and eventually building something useful with these technologies.
Still learning, still experimenting, and sharing the journey. 🚀
Loading article