You are not hiring an engineer. You are hiring for a stack.
A generic technical interview tells you someone can code. It does not tell you whether they can keep a Node event loop unblocked or stop a React tree re-rendering. Each role page sets out the skills, the questions and the shape of the interview for that stack.
30-day free trial with 75 Hire Credits. Card required, nothing charged for 30 days.
- Five round modes per pipeline
- One readiness score per candidate
- Sandboxed code, really executed
Four things, before you write the job post.
Every role page is built the same way, so you can hold two of them side by side and see exactly where the evaluation diverges.
The skill map.
The six things that separate a strong hire from an average one in that stack. You pick these skills and set a weightage on each before anyone runs a round.
The questions we probe.
Real technical questions for the role, not trivia. You can see what a candidate will be pushed on before you invite anyone.
The interview shape.
The four stages we would run for that role, from screening through to the final fit conversation, and what each one is for.
The market band.
An indicative salary range, so the level you are interviewing for and the level you are budgeting for match.
Six roles we have already mapped.
These are the stacks with a ready-made skill map and question set. Any other role works too: you name the skills, set a weightage on each, and the scorecard follows.
What actually changes when the role changes.
Plenty of tools claim a role-specific interview and then ask everyone the same four questions. Four things move when you switch stacks. One thing does not.
- 01
The named skills on the reportcard.
A reportcard is a set of named skills scored from the transcript and the submitted code. You pick those skills and set a weightage on each. For a Node role they are the event loop, streams and memory behaviour. For React they are reconciliation, render cost and state boundaries. A frontend reportcard carries accessibility and page speed, which never appear on a backend one.
- 02
Where the coding round is pointed.
A coding test is a sandboxed problem with real execution against hidden test cases, scored on correctness, efficiency and code quality. What changes per role is the difficulty and the topic tags on the problem. Backend and Java rounds usually read the complexity verdict and the runtime on each test case first. Frontend rounds care more that the logic is right and readable.
- 03
What the interviewer digs into.
The builder has a Questions step, and that is where the role shows up in the conversation. A Node candidate gets pushed on backpressure in a stream pipeline. A Java candidate gets asked to design a thread-safe singleton. The questions adapt to what the candidate says, so depth gets probed where the role needs it.
- 04
The shape of the pipeline.
Most roles run four stages, but not the same four. A React pipeline puts a build-a-component challenge before architecture. A backend pipeline puts API design before system architecture. You assemble it from the same five round modes either way, and AI and human rounds can sit next to each other.
- One readiness score out of 100, scored against the same skills and the same JD for every candidate in the role
- Every score traceable to the transcript, the recording and the submitted code
- Integrity signals surfaced to a human reviewer, never an automatic pass or fail
- A share-ready reportcard with skill breakdown, tags and the round it came from
The same four moves, whatever the stack.
Pick the role page for the detail, then run it. The only role-specific part is what you put into it.
Set the role and the bar.
Pick the skills that matter for this stack, set a weightage on each one, and set the promotion threshold. Set up the JD too, because answers are scored against both.
Structured interviewsScreen the pile against it.
Bulk-upload resumes and run a screening round. Skills, experience and education are read against the role and come back as a ranked table you can sort, shortlist from and export. 1 credit a scan.
Resume screeningRun the technical round.
A coding test with real execution for stack depth, an AI voice round for reasoning and communication, or both in the same pipeline. Add proctoring where the round is high stakes.
Coding testsRead the evidence and decide.
One reportcard per candidate: score, skill breakdown, recording, searchable transcript and malpractice status. Share it or download the PDF. Reports are never metered, and the decision is yours.
How AI rounds workPreparing for one of these roles rather than hiring for it? The question bank and the job seeker tools are built for the other side of the table.
Questions about role-based hiring.
Build the round for the role you are hiring.
Pick the stack, pick the skills and set a weightage, run one round and read the reportcard. Judge us on the evidence, not on this page.
30-day free trial with 75 Hire Credits. Card required, nothing charged for 30 days.
Hiring across several stacks at once? The solutions index covers the pipeline, and pricing shows what each round type costs in credits.