The business says “we want an agent”. Where do you start?
Answer out loud or in your head first, then compare.
Answer
I ask what problem it should solve and what the work looks like today. I want to see a few real examples: what takes time, where errors occur and who checks the result. Only then do I consider the technology.
Plain automation or a single model call may be enough, rather than an autonomous agent. I establish what the system should do, what it must not do and how we'll recognise an improvement. The first experiment should test the biggest unknown, e.g. the reading quality for one type of document. Value doesn't come from simply using AI. The goal is a simpler or more effective process for the user.
Why it matters: A clear way to show independence and product ownership.
Possible follow-ups: What if a quick demo is needed, but criteria for correctness are missing?
How did it go? Be honest — “Not yet” puts the question on your review list.
Start with
- B-P14 · How do unit, integration and end-to-end tests differ? — JavaScript, TypeScript and Node.js
- I-P04 · How does a coding agent differ from a chat that suggests code? — Harness, MCP and coding agents
Further reading
- Official job offer: Senior AI Engineer, RAS Tech(opens in a new tab)Job offer; accessed 22.10.2026. The full scope of the role, stack and integrations. Doesn't describe the exact interview task.