Description
Got tired of seeing roadmaps that are just 300 vendor logos or people claiming you can learn DevOps + ML infrastructure in 8 weeks from YouTube tutorials.
Spent the last few days putting together a text-only, doc-first curriculum. No video playlists, just canonical books (OSTEP, DDIA, ISLP, Beej) and official documentation.
The core idea is you build and evolve a single service—starts as a raw process on a Linux VM, moves into Docker, provisions AWS via Terraform, adds PyTorch training + MLflow lineage, and ends on Kubernetes with OpenTelemetry. Every stage has failure drills where you intentionally break it to learn how to debug.
Paced it for 12–24 months (\~8-10 hrs/week) because doing this while working a day job takes real time if you don't want to just copy-paste without understanding anything.
Site (static, free, no ads/signups): [[link]
Repo: [[link]
Curious what you'd change or what book/doc recommendations I should swap in.
Employer contacts (email/phone/telegram) are hidden from the public preview —
send your CV, and we will connect you directly.