Skip to contentSkip to footer

Ai Software Engineer Interview Questions

Salaries
Interviews

6,031 ai software engineer interview questions shared by candidates

HR round: normal questions to check job fit and culture fit Technical Round: Could you give me a very brief overview about your experience before we dive into the technical questions? Maybe you can tell me about the time where you used either LLM-based reasoning in a system or agentic tools — what you built there, and what you thought was the hardest part about it? Was this already, like, an agentic chat, or was it purely LLM-based without any tools yet? How did the LLM interact with the database? Was the query translating to SQL, or did you use authentic tools there? If we have a situation where you have an ever-growing list of tools and just one agent, how can you deal with that situation? Because at some point, it becomes slower in reasoning which tool to address — how can we come up with something to address that situation? how do you evaluate the system to ensure it was working correctly? Are you familiar with frameworks like LangChain, Pydantic, etc.? What do you think are the benefits of LangChain, and what are maybe some parts that you don’t think work well in LangChain? How about debugging an agentic system — what’s your strategy in debugging? What’s a recent paper, blog post, or technical development in the AI engineering era that you have read about and found interesting?

Senior agentic AI engineer

Interviewed at acto

Nov 11, 2025

HR round: normal questions to check job fit and culture fit Technical Round: Could you give me a very brief overview about your experience before we dive into the technical questions? Maybe you can tell me about the time where you used either LLM-based reasoning in a system or agentic tools — what you built there, and what you thought was the hardest part about it? Was this already, like, an agentic chat, or was it purely LLM-based without any tools yet? How did the LLM interact with the database? Was the query translating to SQL, or did you use authentic tools there? If we have a situation where you have an ever-growing list of tools and just one agent, how can you deal with that situation? Because at some point, it becomes slower in reasoning which tool to address — how can we come up with something to address that situation? how do you evaluate the system to ensure it was working correctly? Are you familiar with frameworks like LangChain, Pydantic, etc.? What do you think are the benefits of LangChain, and what are maybe some parts that you don’t think work well in LangChain? How about debugging an agentic system — what’s your strategy in debugging? What’s a recent paper, blog post, or technical development in the AI engineering era that you have read about and found interesting?

Viewing 5361 - 5370 interview questions

Glassdoor has 6,031 interview questions and reports from Ai software engineer interviews. Prepare for your interview. Get hired. Love your job.