Start here¶
This site is a living learning system, not a static reading list. Three things keep it moving:
- A generated plan. Every day has concrete tasks (
data/curriculum.yml→data/plan.yml→ Today and the week pages). - Daily automation. At 06:00 IST a GitHub Action rebuilds Today, pulls ~30 RSS feeds into the feed, and opens a "Day N" GitHub issue with your checklist. At ~06:15 IST a scheduled Claude agent writes the daily digest: 3–5 must-reads matched to today's topic, a question of the day, and any release that changes something on the site.
- A weekly refresh. On Sundays an agent re-checks versions and links, updates the tech radar, adds new questions, and rebalances next week if you slipped.
Read in this order¶
| # | Page | Why | Time |
|---|---|---|---|
| 1 | The brief | What we're optimising for, constraints, success criteria | 5 min |
| 2 | How to use this site | Daily loop, issues → progress, spaced repetition, question levels | 10 min |
| 3 | Weekly rhythm | Where each track's hours go | 3 min |
| 4 | Phases & gates | What "done" means for each phase | 5 min |
| 5 | Week 0 baseline | Measure yourself before starting | 3–4 h, this weekend |
| 6 | Research: how others build roadmaps | What this roadmap borrows and fixes | 10 min |
| 7 | Freshness contract | How content avoids going stale | 3 min |
The one-screen summary¶
- Goal: Staff/Principal engineer + AI architect who ships production agentic systems. Interview-ready by week 24, measured at 6 checkpoints on one rubric.
- Time: ~17–18 h/week: 12–15 h technical plus 30 min every day of English & communication practice. Weekdays about 2 h, Saturday about 3.5 h (build), Sunday about 3 h (build + review).
- Tracks: Agentic AI (4 h) · System design + AI SD (2.5 h) · DSA (2.5 h) · Architecture (1.5 h) · Python (1.5 h) · Java/Spring AI (1 h) · Staff+ (0.5 h) · Communication & English (3.5 h) · Review (1 h).
- Output: the Agentic Ops Copilot capstone, 6 phase projects, 15+ ADRs, 8 Staff artifacts, 2+ blog posts, and about 250 DSA problems.