No Internships. No Top Papers. Non-CS PhD. My ML Job Hunt in US
This year’s job market has been brutal.
As a non-CS PhD with:
* no top conference publications
* zero internship experience
I sent out countless applications… and got single-digit interviews.
Here’s what the process actually looked like:
TikTok — MLE
* 1 LeetCode
* 2 ML fundamentals
* 2 HM rounds (CV deep dive + LLM)
⸻
McKinsey & Company — Data Scientist II
* Online coding (LeetCode + tabular + sklearn)
* Multiple ML basics rounds
⸻
Meta
Research Scientist – ML
* Phone screen: 2 LeetCode
* Onsite:
* 4 LeetCode
* ML design
* Behavioral
Image Processing / ML Engineer
* Submitted → ghosted
⸻
Patreon — DS transition
* 5-min project presentation (notebook analysis)
⸻
Fish Welfare Initiative
* Google Earth Engine
* Tabular + sklearn coding
⸻
Google — SWE (ML)
* Screening:
* LeetCode + BQ
* Onsite:
* 2 LeetCode
* ML system design
⸻
Reinforce Labs — Applied Scientist
* CTO chat
* ML coding (KNN)
⸻
WeRide — CV Engineer
* CV deep dive + coding
Ask me anything if you want to learn more about my experience.
Navigating the machine learning (ML) job market as a non-CS PhD without the traditional credentials such as internships or prestigious publications can be daunting. From personal experience, the journey requires persistence, strategic preparation, and a focus on demonstrating practical skills. When I first started my job search, I realized how invaluable consistent coding practice is—especially on platforms like LeetCode. Even though I lacked formal internships, mastering data structures, algorithms, and ML fundamentals helped me secure coding interview rounds at top companies. For example, TikTok’s MLE role involved multiple interview rounds focusing on foundational ML and CV deep dives, which pushed me to highlight my technical understanding beyond my academic background. Besides coding skills, I found preparing for machine learning system design was equally critical, especially for roles at Google and Meta. System design interviews assess your ability to architect scalable ML solutions, so I invested time in studying classic models and scalability challenges. Behavioral interviews also played a significant role, giving me an opportunity to communicate my problem-solving approach clearly. Another key takeaway is the importance of tailoring your application materials for each role. Detailed resume targeting—emphasizing relevant projects, coding experience with tools like sklearn, and showcasing any practical involvement such as contributing to initiatives like the Fish Welfare project using Google Earth Engine—helps demonstrate applied expertise. Networking and engaging in community support channels or forums can also uncover hidden opportunities and provide moral support during the lengthy hiring process. Even with rejections or silence (ghosting), persistence kept me moving forward. In summary, your unique background can become your asset if paired with focused skill-building and strategic job application efforts. The 2023 job market may be tough, but a methodical approach to learning and preparation can open doors to ML roles, even without internships or top publications.