My 2024 Wrapped - LinkedIn Edition

Fireworks in the night sky Photo by Ray Hennessy on Unsplash

This year was quite a very interesting one. I learned a lot, did somewhat a bunch not only in Software Development, but also Cyber Engineering, Networking and System Administration.

Although, a lot of the work was for fun and not in a corporate environment, for me, it was still an amazing year to have all this experience acquired in a unique and fun way (late nights debugging because I could not understand why a helm chart does not deploy, and would think about it in my sleep till I get it to work).

Without further ado, here are a list of things I dabbled from this year:

Software Development / DevOps / GitOps

  • Developed a VS Code extension in Typescript that creates/edits H5P files that can be used for interactive assessments for students.
  • Developed a Python application that indexes my Obsidian notes to a Vector database.
  • Migrated away from Docker and Docker-compose and moved to Kubernetes for container orchestration and management.
  • Developed a CI/CD pipeline using GitHub Actions to automate the deployment of Ansible playbooks.
  • Configured an 8-node Kubernetes cluster with 3 masters and 5 workers running rke2 for the homelab.
  • Automated the creation of VMs on Proxmox VE with Terraform and Ansible.
  • Developed an easy and seamless upgrade/update management of the Kubernetes cluster using Ansible and GitHub Actions.
  • Deployed CloudFormation stacks (ECS Fargate Services, VPC, RDS, network configuration and policies) on AWS using the CDK in Typescript.
  • Worked on a Relativity integration for ArkCase FOIA using Java.
  • Learned and did some projects with NextJS.

AI

  • Developed an AI Agent that uses the Vector database for Retrieval Augmented Generation (RAG) using Python.
  • Developed an AI agent that listens to audio, transcribes and replies by text or audio using LLMs (TTS, STT and STS).
  • Developed an AI agent that on request, takes pictures and describes them in detail using LLMs using Python.
  • Developed an AI agent that can generate images using LLMs using Python.

Networking / System Administration / Cyber Engineering

  • This year I have been very proficient in configuring my own OPNSense/PFSense routers, custom DNS providers and ad blockers (AdGuard and Pi-Hole).
  • Configured a home network of 4 segmented networks with one for work, one for the DMZ where I play games to avoid lags, and one for my homelab cluster.
  • Configured firewall rules to restrict access and route traffic between networks.
  • Deployed and configured Identity Providers (IDP) such as Keycloak, Authelia and Authentik.
  • Configured VPNs to connect remotely to my network using Wireguard and Tailscale.
  • Built a computer with 128 GB RAM, 32 cores CPU, 48 GB VRAM (2x24 GB NVIDIA GeForce RTX 4090) that I use to host locally my LLMs.
  • Monitored the performance of my stack with Uptime-Kuma and Grafana and developed dashboards in Grafana to monitor the network status and threats.
  • Automated vulnerability scans and security testing of the homelab stack.

I still have a lot more to talk about on 2024, but I think this will do it as those are on a high level what I could remember.

What’s Next: Looking Forward to 2025

Here are a few things I have pending and that need to be improved in 2025:

  • A fully fledged automated backup system of storage, VMs, databases, etc.
  • Security is an ongoing process, so improve as much as I can in that regard because in 2024, I filtered a lot of what goes out to the internet, my IPS and services providers that use my data. I need to improve that.
  • Write articles about the cool things I work on, as I did not write any in 2024. I need to make time for those.
  • Finish or even get close to a state of which I’m happy with Jeff… (You wonder who Jeff is, huh?). Simply put, he’s an AI assistant that will be an extension of me and help improve my productivity and QOL in many ways. It already encompasses some of the AI work I described earlier.
  • I am excited by the state of AI and open LLMs like Llama 3, Deepseek v3, which I use daily. So I want to explore more on that front as well.
  • Lastly, I need to improve my physical health, which is not bad but could be better.

On this I would say, thank you guys for reading my Ted Talk… oops, I meant LinkedIn Wrap, and I wish you all the best for 2025, and I’m excited about the amazing things you will accomplish as well.