TL;DR: Only ~2% of AI product postings are junior (Axial Search analysis of 12,397 US postings, 2026), so this is a lateral-move market, not an entry-level one. Nobody hands you the title; you arrive with evidence you've already made judgment calls on probabilistic products. This section is the operating manual: path → evidence → targets → loop → level.
| File | What's inside | Use it when |
|---|---|---|
| transition-guides.md | Four paths in (traditional PM, engineer/MLE, data scientist, domain expert), each with its bridge, unfair-advantage roles, and honest timeline | Start here. Pick your path before anything else |
| resume-portfolio.md | Bullet formula, AI-era resume specifics, the portfolio structure (PRD + eval artifact), LinkedIn, the 60-second narrative | You have evidence and need to package it |
| companies-hiring.md | The verified hiring map, by tier: frontier labs, big tech, AI-natives, India & global, with demand data (verified July 2026) | Building your 10-15 company target list |
| compensation.md | Posted ranges and levels.fyi figures for US and India, plus AI-specific negotiation notes (verified July 2026) | Before the recruiter screen, and again at offer |
flowchart TD
A["1. Pick your transition path<br/>PM · engineer · data scientist · domain expert:<br/>each has a different bridge and a<br/>different set of winnable roles"] --> B["2. Build portfolio evidence<br/>shipped projects, one excellent AI PRD,<br/>one eval artifact:<br/>the things almost no candidate has"]
B --> C["3. Target companies by archetype fit<br/>play your background's advantage:<br/>platform/API · feature · AI-native · ML-systems<br/>+ your domain edge"]
C --> D["4. Prep the loop<br/>product sense with AI · technical judgment ·<br/>evals · behavioral, company-specific"]
D --> E["5. Negotiate level<br/>level is the negotiation:<br/>argue scope with evidence, not enthusiasm"]
B -.->|"evidence feeds the answers"| D
Dependencies (this section doesn't stand alone):
- 🛠️ ../04-build-with-ai/projects.md is where the portfolio evidence actually comes from. Step 2 is not optional; every downstream step is weaker without it.
- 🎯 ../07-interview-prep/ has the loops themselves: processes by company, frameworks, question bank.
- Shipped evidence over certificates. Certificates organize learning and signal commitment; none substitutes for built evidence, and hiring managers consistently weight the project. Three certificates instead of one shipped project is a losing trade. (If you need scaffolding, use the verified course catalog, then go build.)
- Claim AI scope in your current role. Now. Propose the AI feature, run the feasibility spike, write the eval-gated PRD. "PM who shipped an AI feature at a non-AI company" beats "PM who took a course," it's available to you this quarter, and internal transfer to your company's AI team is the highest-probability first AI title. See transition-guides.md.
- Referrals via communities. Warm paths dominate AI-native hiring: Lenny's, Mind the Product, Product School Slack, plus build-in-public posts that recruiters explicitly hunt (communities). Watch source boards (Ashby/Greenhouse/Lever) directly; they're faster than LinkedIn reposts.
The honest framing: the demand tailwind is real (median ~$195K across those 12,397 US postings, financial services the #2 hiring industry after tech), but a market with 2% junior roles rewards preparation, not aspiration. Your edge is being the candidate who brings artifacts to the table.
➡️ Next: transition-guides.md. Find your path, then go build the evidence.