I applied through a recruiter. I interviewed at Proton.ai
Interview
Typical start up type interviews. In the last round just had you answer the question “what’s your life story”… met with multiple managers and team members during the interview process 1:1 or in small groups online.
I applied online. The process took 1 week. I interviewed at Proton.ai in Oct 2025
Interview
Weird vibes all around. I applied to the PMM position and reached out to someone at the company directly on LinkedIn - they seemed very interested and asked we schedule a call directly.
I talked to this person the following week, and the call was very odd. They started asking me some conversational questions, then went immediately into grilling me, no context about the company, their role, etc. I did not learn anything more about the company during the interview.
They asked me weird questions (for example, asking me to criticize my current manager) and gave brief, noncommittal answers to my personalized questions about company growth, fundraising, even daily life of the role.
They told me I'd hear back via email. I sent a thank you email the next morning and never heard back.
The overall vibe I get is that this is an organization run mostly by 20-somethings who are looking to hire "friends" more than professionals. The fact I didn't even get an email back tells me that my time wasn't valued, and I got a strange vibe at the pressing questions asked of me during an initial call. I likely would've turned down an offer if I ever got to that stage.
Interview questions [1]
Question 1
How would you rate your current manager on a scale of 1-10?
What is your superpower?
Why did you leave X company?
I applied online. The process took 3 weeks. I interviewed at Proton.ai (Hyderābād) in Aug 2025
Interview
Round 1: Technical screening – SQL, Python, PySpark, Airflow. Clear and relevant.
Round 2: Technical deep dive – architecture, pipelines, scaling, APIs. Challenging but fair.
Round 3 (Final): Discussion with senior leadership. This round was unusual – instead of technical or role-specific questions, it was mostly about my personal life, schooling, and previous managers’ opinions. It felt less like an evaluation of skills/fit and more like digging into personal background.
Feedback:
• Technical rounds were solid and relevant to the role.
• Final round, however, was disappointing – questions were highly personal, not related to data engineering or even cultural fit in a professional sense.
• While I answered everything openly, I felt the interviewers weren’t engaged in listening and that my time was wasted after investing 3 weeks in the process.
Advice to Candidates:
Be prepared that the final round may not be technical and could include deeply personal questions. Decide in advance how comfortable you are with that.
Advice to Company:
I’d encourage Proton.ai to keep the final round focused on skills, cultural fit, and professional experiences. Personal history and relationships with managers aren’t necessary to assess candidate fit and can make the process feel invasive.
Overall Experience: Negative
Interview questions [1]
Question 1
🔹 Technical Rounds
Python & Data Handling
1. How would you read and process an 8 GB text file in Python without running out of memory?
Pipeline Design & Reliability
2. Imagine your pipeline ingests hundreds of CSVs daily from multiple sources. Some arrive late or incomplete — how would you design the pipeline to handle missing/delayed files?
3. Why use Redis checkpointing instead of relying only on Airflow states?
4. What if some files (e.g., products) are missing but others (e.g., orders) arrive — would you still process?
5. What if a file never arrives — how would you handle it?
6. If a late file arrives and the source team asks you to reprocess, how would you re-run the DAG?
7. Are you using Airflow sensors or schedules? How do you pick up files reliably?
APIs & Real-Time
8. If an API experiences downtime, timeouts, or rate limits, how would you design the pipeline to handle it gracefully?
9. Imagine an e-commerce system with 100K products from multiple vendors, prices coming from APIs — how would you design an optimal, resilient system?
10. Would you call APIs on every page load? If not, where would you store and serve the data?
Batch Processing & Optimization
11. Suppose your batch job processing large volumes overnight is too slow — what optimizations would you apply?
12. If the processing is optimized but storage writes are still slow, what would you check/optimize?
13. If instead of Parquet files you had to store in SQL or Elasticsearch, how would you optimize writes?
14. You mentioned indexing — how does indexing affect database writes?
SQL
15. Write a query: Rank sales reps by total sales per region per month using RANK() and GROUP BY.
⸻
🔹 Final Round (Behavioral / Leadership)
16. Tell us your life story starting from high school.
17. What kind of student were you? Which classes did you enjoy or dislike?
18. Back in high school, if we asked what you wanted to become, what would you have said?
19. You played cricket — did you play only casually or for your school?
20. Was choosing your university a deliberate choice? Why that one?
21. How did you do in the entrance exam and what was your thought process in choosing engineering?
22. Apart from studying, what else was going on in your life during college?
23. Who was your first manager? If we call him, what would he say is your superpower?
24. Same question for Gopi (healthcare project manager) — what feedback would he give about you?
25. From all the projects you’ve done, which one are you the most proud of where you had maximum ownership?
26. Why are you looking to leave Newfold / Aeries and what are you looking for next?