Month 12-Week 1: Job-market reconnaissance OR sustained-build mode¶
Week summary¶
- Goal: Decide your post-Q4 trajectory: job search, internal promotion, or continued building. If job-searching: reconnaissance week.
- Time: ~9 h over 3 sessions.
- Output: Map of 30 companies + 10 high-fit roles + tailored materials + 5 coffees scheduled (job-search path); or year-2 acceleration plan (sustained-build path).
Why this week matters¶
A targeted job search outperforms broad applications 10:1. Reconnaissance is the work that makes search efficient. If staying put or continuing to build, this week is for sharpening year-2.
Prerequisites¶
- M11 complete with capstone v0.2 launched.
Recommended cadence¶
- Session A-Tue/Wed evening (~3 h): map the market
- Session B-Sat morning (~3.5 h): identify roles + tailor materials
- Session C-Sun afternoon (~2.5 h): coffees > applications
Session A-Map the market¶
Goal: 30 companies hiring in your specialty. Read 2-3 engineering blog posts each from the top 10.
Part 1-Source companies (60 min)¶
Search: - LinkedIn jobs filter by your specialty terms. - otta.com. - wellfound.com (formerly AngelList). - Discords / Slacks for AI engineers. - Your network.
Capture: name, stage, what they do, why they might want you, public engineering output.
Part 2-Read engineering blogs (75 min)¶
For top 10 by fit: read 2-3 of their posts each. Note: - Vocabulary they use for the role you'd want. - What they brag about technically. - Tone (academic vs scrappy vs corporate).
Part 3-Score fit (45 min)¶
For each of 30: 1-5 score on: - Specialty match. - Stage / scale fit. - Comp likely range. - Geographic fit (remote ok? location ok?).
Top 10 by total score = priority list.
Output of Session A¶
- 30-company map.
- 10 priority companies.
Session B-Identify roles + tailor materials¶
Goal: Find specific open roles. Tailor resume per top-3 targets.
Part 1-Open roles (60 min)¶
For each priority company: find specific open roles. Capture: - Title + link. - 3 bullet points from JD that match your background. - Likely interview format (research on Glassdoor, levels.fyi, Reddit, Levels Discord).
Part 2-Tailor resume per top-3 (90 min)¶
Three resume variants: - Each leads with the project most relevant to that company. - Bullet points reordered to match JD vocabulary. - "Selected Public Artifacts" customized.
Part 3-Cover letter template (30 min)¶
ONE per-target cover letter is fine. Template:
[Specific reason this company]
I'm applying for <role>. My background:
- <Bullet 1: specialty>
- <Bullet 2: bridge-backend/observability>
- <Bullet 3: shipped capstone with measurable result>
Selected work:
- <link>
- <link>
I'd love to talk about <specific question about their work>.
Output of Session B¶
- 10 high-fit roles identified.
- Resume variants for top-3.
- Cover letter template.
Session C-Coffees > applications¶
Goal: Schedule conversations with people at target companies. Do not cold-apply yet.
Part 1-Identify warm paths (60 min)¶
For each priority company: who do you know (or know of) who works there? - Direct network (LinkedIn 1st-degree). - Network of network (2nd-degree, ask for intro). - Public technical contacts (people who blog, talk, contribute OSS). - People you've already reached out to in M11-W03.
Part 2-Schedule 5 coffees (60 min)¶
Reach out to 5 specific people. Not for jobs-for technical conversations about their work.
Hi <name>,
I'm in the middle of a career transition into AI engineering with a specialty in <X>.
I built <capstone link> and wrote about it here: <post>.
I'm exploring what teams like <company> are doing in this space. Would you have
20 minutes for a video chat in the next 2 weeks?
Part 3-Read about target interview formats (30 min)¶
For top-3 targets: search "
Common 2026 formats: - Algorithmic coding (LeetCode-medium). - ML system design. - ML breadth (transformers, evals, inference, RAG basics). - Take-home (build small LLM app over a weekend). - Behavioral / values.
Note formats per target. Ground for next week's interview prep.
Output of Session C¶
- 5 coffee asks sent.
- Interview format research per top-3 target.
End-of-week artifact¶
- 30-company map
- 10 priority roles identified
- Resumes tailored per top-3
- 5 coffees scheduled or pending
- Interview format research
End-of-week self-assessment¶
- I know specifically who I want to work with.
- I know roughly what their interviews look like.
- I have warm-path conversations starting.
Common failure modes for this week¶
- Cold applications first. They have 1-2% response rates. Coffees first.
- Generic outreach. Each message specific.
- Skipping format research. Surprise format = bad interview.
What's next (preview of M12-W02)¶
Interview prep. Coding refresh, ML system design practice, breadth review, behavioral stories.