Company Research & Career Growth

Researching Companies

Before Interview:

1. ML Tech Stack (ask in recruiter call):

   • What ML frameworks? (PyTorch, TensorFlow, JAX)
   • Cloud provider? (AWS SageMaker, GCP Vertex, Azure ML)
   • MLOps tools? (MLflow, Kubeflow, Weights & Biases)
   • Production serving? (TensorFlow Serving, TorchServe, custom)

2. ML Team Structure:

   • Team size? (2-person vs 50-person team)
   • Reporting structure? (ML-specific manager vs product manager)
   • Collaboration with research team?
   • On-call expectations?

3. Data & Problems:

   • What data scale? (GBs, TBs, PBs)
   • Problem types? (recommendation, NLP, CV, forecasting)
   • Business impact of ML? (core product vs supporting feature)

4. Culture Research:

   • Blind app: Real employee discussions
   • Glassdoor: Reviews (filter for ML roles)
   • LinkedIn: Message current ML engineers
   • Engineering blog: Read recent posts

Questions to Ask Interviewer:

"What's the biggest ML challenge your team is tackling?" "How do you measure success for ML projects?" "What's your model deployment process?" "How often do models get retrained?" "Tell me about a recent ML project that failed and what you learned"

Career Growth Paths

IC (Individual Contributor) Track:

Junior ML Engineer (L3) → ML Engineer (L4) → Senior ML Engineer (L5) → Staff ML Engineer (L6) → Principal/Distinguished (L7+)

Timeline:

• L3 → L4: 1-2 years (demonstrate independence)
• L4 → L5: 2-4 years (lead projects, mentor)
• L5 → L6: 3-5 years (org-level impact, technical leadership)
• L6 → L7: 5+ years (company-wide impact, industry recognition)

What Each Level Requires:

L4 (ML Engineer):

• Own end-to-end ML projects
• Debug production models independently
• Write design docs for medium projects

L5 (Senior):

• Lead cross-team projects
• Mentor 2-3 junior engineers
• Drive technical decisions
• On-call rotation ownership

L6 (Staff):

• Set technical direction for organization
• Influence roadmap across teams
• Mentor senior engineers
• Represent company externally (conferences, papers)

Management Track:

ML Engineer → ML Manager → Sr. Manager → Director → VP of ML

Switch Point: Usually at Senior level (L5)

• Manager: People leadership, roadmap, hiring, 1:1s
• IC: Deep technical, architecture, no direct reports

Building Your Career

0-2 Years: Foundations

• Master 1-2 ML domains deeply (NLP, CV, RecSys)
• Ship 3-5 production models
• Contribute to open source
• Build portfolio projects

2-5 Years: Specialization & Leadership

• Become go-to expert in one area
• Lead 2-3 junior engineers
• Publish blog posts/talks
• Start building industry network

5+ Years: Impact & Influence

• Publish papers at conferences (NeurIPS, ICML, ICLR)
• Build OSS tools used by others
• Tech talks at conferences
• Consider: Startup, research lab, or big tech principal track

Continuous Learning:

• Read 2-3 papers/week (arxiv.org)
• Implement papers from scratch
• Contribute to OSS (PyTorch, Hugging Face)
• Side projects with new tech
• Teach others (blog, talks)

Switching Companies:

• Every 2-4 years for salary bumps (20-40% raises)
• Stay 2+ years to vest equity
• Don't job-hop too frequently (red flag)

:::