After Level 5 the roadmap splits into four career tracks: AI App Developer, AI Engineer, ML Researcher and AI Product Manager. Each takes 8 to 14 weeks and ends with four portfolio projects aimed at one role.
By the Build Fast with AI teamPublished Updated
Compare
| Path | Pick it if you want to | Duration | Target role |
|---|---|---|---|
| AI App Developer | Ship AI products people use every day | 8-10 weeks | Senior AI/Full-Stack Engineer |
| AI Engineer | Build AI systems that stay fast, cheap and reliable at scale | 8-12 weeks | ML/AI Systems Architect |
| ML Researcher | Invent, test and publish new methods | 10-14 weeks | ML Research Scientist |
| AI Product Manager | Decide what AI features to build and how to measure them | 8-10 weeks | Product Manager / AI Product Lead |
Tracks
Specialize in building consumer-facing GenAI applications. Master frontend frameworks, real-time communication, and product-focused development.
8-10 weeks ยท 4 projects ยท Senior AI/Full-Stack Engineer
Specialize in building scalable, efficient AI systems. Focus on system design, optimization, and infrastructure.
8-12 weeks ยท 4 projects ยท ML/AI Systems Architect
Run original experiments, publish papers and reproduce the newest methods in retrieval, agents and fine-tuning.
10-14 weeks ยท 4 projects ยท ML Research Scientist
Lead AI products and strategy. Master customer development, metrics, and go-to-market strategy.
8-10 weeks ยท 4 projects ยท Product Manager / AI Product Lead
Details
Open a path to see its core skills and the four portfolio projects.
8-10 weeks ยท Senior AI/Full-Stack Engineer
8-12 weeks ยท ML/AI Systems Architect
10-14 weeks ยท ML Research Scientist
8-10 weeks ยท Product Manager / AI Product Lead
FAQ
Yes. Levels 0 to 5 are shared by all four tracks, so switching only costs you the track-specific projects. Commit to one path for at least 8 weeks before deciding it is the wrong fit.
We recommend finishing Level 5 first. If you already build LLM apps with RAG and have deployed one to production, you can start a track earlier and fill gaps as you go.
Yes. Many people pair AI Engineer with AI App Developer, or AI Product Manager with either engineering track. Plan for 16 to 28 extra weeks if you take two tracks back to back.
For most developers it is AI App Developer or AI Engineer, because both reuse skills you already have and the portfolio projects map directly to job descriptions. ML Researcher usually takes longest because publications and reproductions take time.
Every path assumes Levels 0 to 5. Start at the beginning, or jump to the first level that covers something you can't do yet.
See the full roadmap โ