Technical assessment
Not every technical role is a coding role. Dextes lets you assess cloud, data, security, quality and platform skills with the same tools you would use to screen a developer, and gives you a report that says which skills a candidate actually has.
No credit card required
Most screening ends with one number. Seventy-two per cent. It sorts candidates into an order, and then stops being useful - because the two people on either side of your cut-off may have got there in completely different ways.
For a technical role that matters more than usual. A cloud engineer who is strong on networking and weak on security is a different hire from one with the reverse profile, and both of them can score seventy-two. The score hides exactly the thing you needed to know.
An assessment is only worth running if the result tells you something specific enough to act on.
Dextes ships with 5,990 skills organised into 52 categories, 27 of which are technology categories. You tag questions with them, and the tags are what make a per-skill report possible.
AWS, Microsoft, Google, Alibaba and VMware, as separate categories rather than one "cloud" bucket.
Data platforms, analytics and machine-learning skills.
Security skills as their own category, not folded into infrastructure.
SAP, Salesforce, Oracle, Workday, Adobe, IBM and Pegasystems.
Testing and quality skills for QA and SDET roles.
Delivery practices, from Scrum through to platform engineering.
System and solution architecture, including legacy decoupling.
Application development, mobility, and emerging areas like blockchain and extended reality.
The library also covers 20 industries and the non-technical business functions, so the same account can assess a procurement analyst or a recruiter without a second tool.
Answers are grouped by the skills their questions were tagged with. For each skill you get how many questions were answered correctly and how long they took on average.
Skills performance — example
| Skill | Correct | Avg time |
|---|---|---|
| Amazon Web Services | 7 / 8 | 41s |
| Kubernetes | 6 / 8 | 58s |
| Network security | 2 / 6 | 1m 34s |
| Infrastructure as code | 5 / 6 | 47s |
The figures above illustrate the layout. They are not data from a real candidate.
All three produce an ordinary assessment you can edit afterwards.
Pick the skills the role needs and build questions against them. Because the questions carry those skill tags, the report comes back broken down the same way.
Paste the JD. Dextes extracts the role and the skills it implies, and drafts a first set of questions with a suggested duration and pass mark. Everything is editable before you send it.
Write your own questions, or reuse ones already in your bank. Questions you keep are organised in folders and can be pulled into any later assessment.
Each is its own assessment, so you can send a functional screen first and a cognitive or psychometric one only to the people who get through it.
Role and skill knowledge, the default for most technical screening.
Problems solved in an in-browser editor and run against your test cases.
Reasoning ability, as its own assessment rather than a question type.
Trait-based assessment, scored against a framework rather than right and wrong.
Screening developers specifically, with coding as the main event? The developer assessment page covers the in-browser editor, the supported languages and how submissions are run.
Multiple choice and multi-select are marked the moment an assessment is submitted. Written answers go to an AI model, which scores each one against your reference answer and returns a short explanation of why it gave that score.
The explanation is the point. A score on its own is not a verdict you can defend in a hiring discussion, and any AI score can be overridden. Answers the model could not grade confidently are routed to a review queue rather than given a guessed score.
Assessments can enforce a time limit on the server, record tab switches against the attempt, warn a candidate when it happens, and optionally submit automatically once a threshold is crossed. Copy and paste, right-click and full-screen exit can each be restricted, and links can be set to expire or to work only once.
What there is not: no webcam recording, screen recording or identity verification. If your process requires supervised proctoring, Dextes is not a substitute for it.
Any assessment built around technical skills - cloud platforms, data and AI, security, quality engineering, agile delivery, or enterprise platforms like SAP, Salesforce, Workday and Oracle. Coding questions are available but not required, so you can assess a cloud engineer or a QA lead without asking them to write an algorithm.
Every question carries skill tags, and the candidate report groups answers by skill. You see how many questions were answered correctly in each skill and the average time spent on them, rather than one overall percentage.
No. You can build from the skills library, paste a job description and have the questions drafted for you, or write them yourself. Anything generated is editable before you send it, and it becomes part of your own question bank.
Yes. Multiple choice, multi-select, written answers and coding questions can all appear in the same assessment. Multiple choice and multi-select are graded automatically, written answers are AI-graded with the reasoning shown, and coding submissions run against your test cases.
It is the same platform. The developer assessment page covers coding-heavy screening in detail - the in-browser editor, supported languages and how code is run. This page is about assessing technical skills across roles where code is only part of the job, or none of it.
Build an assessment from the skills library or from a job description, send it, and read the breakdown.