Beyond my day job, I'm diving headfirst into the world of AI and Machine Learning. Discover my journey back to Python, the Udemy courses fueling my learning, and the exciting plans for a mini home lab to run local LLMs.
As full stack engineers, we're constantly juggling new frameworks, evolving cloud services, and shifting architectural patterns. But beyond the day to day, there are always areas that truly spark our curiosity and drive us to learn more. For me, that area is increasingly Artificial Intelligence (AI) and Machine Learning (ML).
The advancements in AI over the past year have been nothing short of breathtaking. From large language models (LLMs) revolutionising how we interact with information to AI driven tools reshaping development workflows, it's clear this isn't just a trend it's a fundamental shift. And as someone who loves to build and understand systems end to end, I'm incredibly keen to move beyond just reading about it to actively learning and experimenting.
Back to Basics: Re-embracing Python
My first step has been a delightful return to Python. While my full-stack work often involves JavaScript/TypeScript on the front end and various back end languages, Python remains the undisputed champion of the AI/ML world. It's been great to refresh my memory on its elegant syntax and rich ecosystem, particularly with libraries like NumPy, Pandas, and Scikit-learn, which are the foundational tools for any data scientist or ML engineer. It feels good to flex those different coding muscles again!
Structured Learning: My Udemy Deep Dive
To get a structured understanding of the theoretical underpinnings and practical applications, I've enrolled in several highly rated courses on Udemy. These courses cover everything from the mathematical concepts behind neural networks to hands on implementation of various machine learning algorithms.
The goal isn't just to passively consume information, but to actively code along, build small projects, and solidify my understanding of concepts like supervised vs. unsupervised learning, model training, evaluation metrics, and the ethical considerations that are so crucial in this field.
The Mini Home Lab: Running LLMs Locally
One of the most exciting aspects of current AI research is the increasing capability of smaller, more efficient LLMs. These models, while not as vast as their cloud based counterparts, are becoming powerful enough to run locally on consumer grade hardware. This has ignited my interest in setting up a mini home lab environment specifically for AI/ML experimentation.
My plan is to start with a basic PC configuration that can handle the computational demands of running these smaller LLMs. The idea is to:
Experiment Freely: Have a dedicated environment where I can download, fine tune (even slightly), and run various open-source LLMs without incurring cloud costs.
Understand Performance: Gain practical experience in optimising models for local inference and understanding the hardware requirements.
Build Personal Projects: Eventually, integrate these local models into personal full stack projects, perhaps creating custom AI agents for specific tasks or automating aspects of my home network.
This journey into AI and ML is more than just learning new tech; it's about staying curious, pushing my boundaries, and understanding the next wave of innovation that will undoubtedly shape the future of every industry, including FinTech. It's an exciting path, and I'm eager to share my progress as I delve deeper.