Frequently Asked Questions
How do I get started with Machine Learning?
Start with the fundamentals: linear algebra, probability & statistics, and Python programming. These are non-negotiable foundations that everything else builds upon.
Once you're comfortable with the basics, move on to understanding classical ML algorithms (linear regression, logistic regression, decision trees, random forests, SVMs, k-means). Don't rush to deep learning - understanding these fundamentals will make everything else click faster.
For deep learning, start with neural network basics, then specialize based on your interests: computer vision (CNNs), NLP (Transformers), or other domains.
Most importantly: build projects. Theory only gets you so far. Implement papers, participate in Kaggle competitions, or build something useful. The gap between knowing ML and doing ML is bridged by practice.
Check out my ML Career Advice section for more detailed guidance.
How do I get into big tech?
No secret here: you need a CV strong enough to get shortlisted, and then you need to prep for interviews. Simple, but not easy.
The key insight is this: identify where you're getting stuck in the process, and that's where you optimize.
Not getting callbacks? Work on your CV, build more impressive projects, or get referrals. Failing phone screens? Practice coding problems. Bombing system design? Study distributed systems and practice articulating trade-offs. Struggling with behavioral? Prepare your stories using the STAR method.
Don't spray and pray. Be methodical about diagnosing your bottleneck and fix that specific thing.
What's the difference between MLE and Data Scientist roles?
Machine Learning Engineers focus more on building production systems, infrastructure, and deploying models at scale. Data Scientists typically focus more on analysis, experimentation, and model development. However, the lines are often blurred and vary significantly between companies. I wrote about this in detail: The most overloaded role: "Machine learning engineer".
Should I do a PhD to join big tech?
No, if your only goal is to join big tech. There are plenty of easier ways to get there. A PhD takes 4-6 years of your life, pays significantly less than industry, and isn't required for most roles including MLE and Data Scientist positions.
However, if you want to be a Research Scientist and only that, then yes, you might need a PhD. RS roles at top labs typically require a PhD with a strong publication record in your area of focus.
The key point: don't do a PhD solely with the goal of working in big tech. If you're doing a PhD just to get hired at Google or Meta, you're taking a much longer and harder path than necessary. Do a PhD because you genuinely want to spend years deeply exploring a research problem, not as a job credential.