ServiceNow Internship Experience (QA → ML → ???)
Summary
I interviewed for an Associate Machine Learning Engineer Intern role at ServiceNow, an experience that transitioned from an initial QA internship. The process included a technical screening with DSA questions (Anagrams, Find the Duplicate Number), an unexpected QA/SDET focused round with test case writing, and a confusing technical deep dive requiring system design (product review sentiment analysis) and delving into various Java/JVM and Spring Boot topics, ultimately leading to no selection.
Full Experience
Posting this to share my experience and maybe get some perspective from the community.
Background
- Role initially: QA Intern
- Later changed (via email): Associate Machine Learning Engineer Intern
- OA conducted via college
- 6 candidates shortlisted, 2 reached final rounds
Round 1 – Technical Screening
- Format:
- Quiz-style questions (similar to Google-style logical rounds)
- Company-related questions
- DSA:
- Anagrams
- Find the Duplicate Number (LC 287)
- I Explained brute → optimized approachesm ,Implemented optimal solution,Answered follow-ups ,discussed ML/LLM projects (ML vs GenAI, transformers, etc.)
- Result: Only 2/6 candidates shortlisted
- I cleared this round
Round 2 – HR / Process (Unexpected Direction)
This was supposed to be HR.
First ~20 minutes:
- Project discussion
- Communication, background, experiences
- Then suddenly shifted to:
- Writing test cases for Amazon’s product cart system
- Questions on SDLC
- QA/testing workflows
- This felt more like a QA/SDET round, which was confusing given the role had been changed to ML Engineer Intern.
Round 3 – Technical Deep Dive (Most Confusing)
Asked to design a system:
- User submits a product review
- System determines positive vs negative
- Positive → redirect to thank-you page
- Negative → create ticket + redirect
My approaches:
- Transformer-based semantic embeddings → rejected (“No LLMs”)
- ML-based approach → rejected
Asked to write code
- Constraints:
- Language: Java
- I implemented a basic working solution using weighted negative keywords.
Then the round escalated into:
- Sarcasm handling
- JVM internals
- Memory management
- String vs StringBuilder(SCP)
- Thread safety
- JAR files
- Spring Boot related questions
At this point, it felt less like an intern evaluation and more like a stress test across unrelated areas. [my projects were in mern/pern stack]
Outcome
- After ~1 month:
- No candidate was selected
Posting this mainly to Share experience and Understand if this kind of process is normal.
Would love to hear thoughts from people who’ve interviewed at large companies recently.
Interview Questions (13)
Determine if two given strings are anagrams of each other.
Write test cases for an Amazon-like product cart system.
Questions related to the Software Development Life Cycle (SDLC).
Questions related to QA and testing workflows.
Design a system where a user submits a product review. The system determines if the review is positive or negative. If positive, redirect to a thank-you page. If negative, create a ticket and then redirect. Constraints included language: Java.
Discussion on handling sarcasm in natural language processing.
Questions about Java Virtual Machine (JVM) internals.
Questions on memory management, likely in the context of Java.
Discussion on the differences between String and StringBuilder, and concepts related to the String Constant Pool (SCP).
Questions on thread safety.
Questions about JAR files.
Questions related to Spring Boot framework.