My AI/ML learning journey (so far...)

javitech.co — Sat, 06 Sep 2025 03:55:07 +0000

In the last year I’ve been studying about AI and machine learning. Even when I started using ChatGPT early in 2023, my first deliverable using AI was on September 2024 when I participated in a challenge organized by Platzi and NewRelic , where I had to fix bugs and improve an Angular app (it worth mention that my main role is SRE/DevOps, knowing how to code), using Github Copilot.

A couple months later, I started playing with local tools like ollama, gpt4all and LM Studio. At this point I didn’t knew that much the tools, but I followed the philosophy of learn by doing. It reminds me a little bit when I started learning Docker, where I installed software and followed tutorials even when I didn’t really knew what I was doing, that came later with time after using the tools regularly and reading books to dig deeper in the topic.

After that, in December 2024, I participated in a challenge organized by CodigoFacilito and Microsoft, in which I learned a lot of AI/ML and got Azure AI Engineer Associate Certified. Along with this preparation phase, I worked with Leonel Alberto in a project for 2 weeks, besides putting into practice our knowledge in AI/Python, I think that the most valuable take away of this experience was the communication and organization to deliver not just the software, but also documentation, presentation, etc, as a whole product, it was definitely a team work and I encourage everyone to give it a chance to work in challenges, it’s not just the technical part that is improved on those events but also the behavioral skills.

Having received the Azure AI Engineer certification, I encouraged myself to prepare and take the AWS Certified Machine Learning Engineer - Associate certification, as AWS is the cloud provider that I have more experience, I passed the exam early in 2025. After that I’ve been participating in AWS initiatives like AWS Build Games where I built a retro game from scratch using Amazon Q cli

At this point, feeling more confortable with AI/ML, I took the next step my study journey by reading “Designing Machine Learning Systems” O’reilly book by Chip Huyen, in my opinion a well written and organized book, full of references, advices and personal comments that are invaluables, I’ll definitely keep this book close for future reference.

Along side reading this book, I’ve also been following Stephanee Maarek and Frank Kane Udemy Machine Learning Specialty Course, as well as Antonio Feregrino’s CodigoFacilito material. Apart from having pedagogy, they transmit the experience they have in the subject, recommended.

I’ve learned a lot in the last year (it feels like it has passed more that a single year), and my next goals are:

build another project using machine learning
Read Effective Data Science Infrastructure
Pass AWS Machine Learning Specialty Certification

I would like to hear from you in the comments, how has been your learning journey and adaptation in the world of AI/ML? what recommendations do you have for me or other people in the same learning path? I read you

This is how I created a game from scratch using AWS

javitech.co — Tue, 24 Jun 2025 23:07:13 +0000

A couple of weeks ago I heard about AWS's Build Games Challenge, I got excited as I saw it as an opportunity to build my first game. At that time the challenge wasn't considering latin american countries, so I couldn't participate, but it was a pleasant surprise when I saw an update on the post where AWS was now accepting more countries, including mine, there were no more excuses for me.

As the theme of the challenge was to build retro games, first I started to ask myself, what's "retro"? as for some people it could be a black and white, pixels only Pong game, and for others it could be N64 GoldenEye. I decided to stick with the games that I played as a kid, being nintendo (NES) platform, games like Ninja Gaiden, Megaman, Mario 3 among others really leaved an impression on me, so I started to think which of those games I could build as my first approach to video game development, I recall a mini game on Mario 3, memory card game.

My intention was to create a game that has some relation with AWS, to facilitate the newcomers on AWS world to get used to AWS services, let them know some AWS Servies names, and which ones are fundamentals, and which ones are more advanced.

I was impressed by Amazon Q, first it helped me on the description of the prompt, that was very helpful as I was considering a list of things, but others were out of my sigh, like adding a timer and other little details that I'm sure helped on completing the game faster by being more specific. Also I really like it that Q iterated on it's own proposals, anticipating fine tunes from me, for example, initially it tried to use emojis for the cards, but without me asking, it iterated and make the icons as SVG format to keep the escense of retro game. Also Q identified screenshots of the game on the root of the repository, and it added them as examples in the README.md file, that was very nice.

I discovered that Q support not only text (prompt), but also images, I just had to reference the filename and it handled it perfectly, no special command or tag to make it work, it was very natural.

My role in the past years has been mostly on Site Reliability Engineer side, so I feel confortable with tools like Infrastructure as Code, Continuos Integration, so I leverange on Amazon Q not only to build the game, but also to automate the deployment process, on creating the infrastructure and CI/CD pipeline, it was perfect for my workflow, I knew what I wanted, I asked for it and then I reviewed the proposed code, it was almost a one-shot prompt, that kept me on the flow which is the porpuse of the tools that we use, accomplish goals.

Here are some screenshots of the game 👇
Easy mode 😌: 4 pair of cards ; fundamental AWS Services

Hard mode 😰: 9 pair of cards ; advanced AWS Services

🎮 Play now: http://aws-memory-game-javitech.s3-website-us-east-1.amazonaws.com

✨ Source Code on Github repository: https://github.com/dvst/game-memory-aws-build-challenge

Do you like it the game? give it a star on github. have any ideas to improve the game? play with it and create a Pull Request!

📺 YouTube of video game creation process from scratch (Spanish with English subtitles): https://www.youtube.com/watch?v=ZnrQxi8GPhk

DEV Community: javitech.co

My AI/ML learning journey (so far...)

This is how I created a game from scratch using AWS