Coding. A real editor, and someone asking why.
Forty-five minutes with a Monaco editor and a live interviewer, on a call the whole time. The difference from an async test is not the editor, it is that someone asks about the choice you just made.
Correctness is capped, not decisive
This is the entire argument against async coding tests, and it is a deliberate product position rather than a limitation.
A candidate who fails two hidden cases but reasons precisely about why can outscore one who passes everything and cannot explain the approach.
An automated test cannot tell those two apart, which is why practiced candidates beat capable ones on those platforms. Technical Depth is one of four dimensions the interview scores, each a single 1-to-5 score carrying its own written rationale and cited evidence, never a pass/fail count.
| Dimension | What it is |
|---|---|
| Problem Decomposition | Breaking the problem into the right pieces before writing code |
| Architectural Judgment | The tradeoffs behind the approach, not only that it runs |
| Technical Depth | Correctness, complexity and code quality, judged together as one score |
| Communication | Whether the reasoning was followable as it happened |
The workspace
What it does
- A live conversation alongside the editor: the interviewer asks about the approach as you talk it through, the same way a strong human interviewer would
- A transcript with timestamped evidence citations against the published rubric, which is what the report is built from
- A scratch pad for complexity notes
- The candidate’s preferred language, opened by default, and switching is never remarked on
What it does not do
- Aggressive autocomplete. LSP completion is off by default, because it corrupts the signal you are actually being evaluated on
- Proctoring. No webcam monitoring, no screen locking, no browser lockdown
- Run on a phone. The interview refuses rather than measuring your thumbs
How it is weighted
Each dimension is weighted differently depending on the level being interviewed for, so Technical Depth does not count the same at Junior/Mid as it does at Lead/Staff+. The exact weights for every level are in the table on the rubric page.
Compare it to the test you already own.
Most companies reading this have an async platform. The question is what it can and cannot tell you.
The other interview types: Technical Screen · System Design · Data Science