Mostly on System Design - Machine Learning
I was asked questions on the below topics:
Model deployment
Train Loss evaluation
Eval metrics for LLM
Observability platforms
Data Drift
CI CD such as Github Actions
Canary Deployments
Designing Agentic AI
Use of frameworks vs custom build
RAG vs Fine Tuning ( heavily on use cases )
Interview questions [1]
Question 1
Model Drift on production, its monitoring & alerts & strategies therein
I applied online. I interviewed at Hy.ly in July 2025
Interview
I got on a call with the HR as a first round of interview. The questions were around my current role and experience.
In the end HR asked about current CTC and mentioned that their budget is half of my current salary.
It was a waste of time. The screening could have been done before scheduling a call.
Also, to apply for this role, a google form with so many documents need to be filled which seems unnecessary.
Multiple rounds and different folks took the interviews across 2 weeks
Presentation round to review skills and other things.
Process starts with a linkedin post and screening, then a form, then an HR interview, first round by team and then leadership takes the final round.