What to assess in a React developer
A React role is mostly about state, data flow and the edges of the UI. Syntax matters less than it used to, because an assistant will supply it. Judgement about where state lives, when to fetch, and how to test a component still has to come from the engineer.
- Component design: splitting a screen into components with clear props, and knowing when not to split.
- State and effects: local versus lifted versus server state, effect dependencies, and avoiding needless re-renders.
- Data fetching: loading, error and empty states, race conditions, and cancelling stale requests.
- Accessibility: semantic elements, labels, focus management and keyboard use.
- Testing: writing a Testing Library test that checks behaviour rather than implementation details.
- Working with AI: asking precise questions, reading generated code before accepting it, and testing it.
Screening questions that reveal real React experience
SynergicHire drafts these from your job description, and you edit and approve the plan. The screener asks follow-ups when an answer is vague, so a memorised definition is not enough.
- A list re-renders every time the parent's input changes. How would you find out why, and what would you change?
- When would you keep state in the URL rather than in a component?
- Walk through what happens when a user types quickly into a search box that fetches results. What goes wrong, and how do you prevent it?
- An assistant suggests wrapping every callback in useCallback. Would you accept that? Why or why not?
- How do you test a component that loads data on mount?
The React coding round
The coding round opens a React (Vite) project with starter files, Testing Library tests and a live preview. The candidate works in a Monaco editor with a terminal, inline completions, file-aware chat and an agent. You choose guarded mode, which tests fundamentals with a limited helper, or unguarded mode, which tests how they ship with a full assistant. The server enforces the mode.
Code is tinted by who wrote it: typed by the candidate, or accepted from the AI. When the round ends, the server runs the tests itself. A pass reported by the browser is not trusted.
AI-fluency signals to read
| Signal | What it tells you about a React developer |
|---|---|
| Time to first prompt | Whether they read the components and tests before asking the assistant. |
| Prompt quality | Whether prompts name the component, the props and the failing behaviour, or just say "fix it". |
| AI-generated share | How much of the final code came from accepted AI output. Neither high nor low is good on its own. |
| Verification | Whether they ran the tests or edited the code within 90 seconds of accepting a suggestion. |
| Over-reliance flags | Large blocks accepted and never reviewed or tested, such as a whole hook pasted in unread. |
| Recovery | What they did when the assistant was wrong: fixed it, re-prompted with context, or moved on. |
From scores to a decision
Fundamentals and AI fluency are scored separately and never averaged together. A score that cannot cite a real answer or a real edit is dropped and reported as a gap. The recommendation is advice, such as "Worth a technical round". Nothing advances or rejects a candidate on its own: a named person signs off every round and records every outcome.
Questions
Can candidates use AI in a React interview?
Yes, if you choose. Guarded mode gives a limited helper and tests fundamentals. Unguarded mode gives a full assistant and tests how they build with it. The server enforces the mode, so a candidate cannot unlock more help by pasting instructions into the chat.
Does the React coding round run real code?
Yes. Each round runs a real React (Vite) project in a sandbox with a terminal, tests and a live preview. The server reruns the tests when the round ends.
How long is a React coding round?
You set the length. The price of a coding round covers up to 90 minutes. The clock runs from the start and does not pause.
Is SynergicHire only for React?
No. Coding rounds also run Next.js, React Native (Expo) and Node.js projects, and the AI screen can cover any competency in your plan. Hands-on code in the screen is TypeScript.