Features
One platform, from question to hiring decision
Everything below is in the product today. Where a capability does not exist, we say so on this page rather than letting you find out during a trial.
Authoring
Build the assessment
A question bank you can grow and reuse, not a one-off form per assessment.
Nine question types
Objective types for knowledge, coding problems for ability, and long-form types for reasoning you want to read yourself.
A bank that stays navigable
Tags, categories and difficulty on every question, with search across the body. Reuse a question in any number of assessments.
Randomising pools
Define a pool and how many questions to draw. Each candidate gets a different paper from the same specification, so sharing answers between slots does not help.
Sections with their own rules
Split an assessment into MCQ, coding or mixed sections, each with its own question set and weighting.
Timing and login windows
Set the duration, and optionally the window during which candidates may start. A drive closes on schedule without anyone watching the clock.
Review before you publish
Questions move through draft, review and published states, so an untested question does not reach a candidate by accident.
Scored automatically
- Single choice
- Multiple choice
- Fill in the blank
- Image based
- Code snippet MCQ
- Coding problem
Graded the moment an attempt is submitted, including coding against your test cases.
Scored by a reviewer
- Paragraph / comprehension
- Subjective / long answer
- Case study
Prose is not auto-scored. These go to a review queue with your rubric.
Delivery
Run it, at whatever scale you need
From a single interview to a campus drive with thousands of candidates in one window.
Full code editor
Syntax highlighting, autocomplete and a layout the candidate can arrange. Answers and code save as they type, so a dropped connection does not lose work.
Sandboxed execution
Code runs against your test cases in an isolated container with enforced CPU and memory limits. Candidates can run before they submit.
Identity verification
Optionally require a photo ID and a live capture before the assessment starts, with the comparison recorded against the attempt.
Pre-flight system check
Browser, camera, microphone, screen share, connection speed and single-display setup are verified before the timer starts, with plain instructions for anything that fails.
Live invigilation
See every active attempt with progress and flags as they happen. Message a candidate, pause them, extend time, or force-submit when a session has to end.
Candidate email
Invitations, reminders and result notifications are sent for you, carrying your organisation's name and branding rather than ours.
Execution languages
- C
- C++
- Python
- Java
- JavaScript
- TypeScript
- Go
- Rust
- Kotlin
- Swift
- PHP
- Ruby
- SQL
Each runs in a sandboxed container with its own time and memory ceiling. SQL problems execute against SQLite and are checked on the output the query produces.
Integrity
Evidence, not just a verdict
A risk score is only useful if you can see what produced it. Every signal is retained against the attempt.
Webcam and screen monitoring
Periodic snapshots from both, retained against the attempt so a flag can be checked rather than taken on faith.
Weighted risk score
A 0–100 score per attempt with a low, medium or high band. Every point traces back to the specific events that caused it.
Cohort plagiarism detection
Code compared across candidates by token sequence, syntax tree and structure, with matching regions shown side by side for a human to judge.
Attempt timeline
Every event in order — when a tab lost focus, when a face left frame, when a paste happened — so an integrity question has an answer.
Review
Decide, then defend the decision
Results that a hiring manager can read and an auditor can trace.
Result analytics
Score distributions across the cohort, section breakdowns and per-question difficulty, so you can tell a hard question from a bad one.
Manual review queue
Long-form answers land in a queue with your rubric alongside. The result records who scored it and when.
Export
Results to PDF for sharing and Excel for your own analysis, including integrity scores and section breakdowns.
Roles and scoping
Eight roles across 22 permissions, per-member overrides, and assessment-level access so an interviewer sees only their own drives.
Audit trail
Privileged actions are recorded with who did them, to what and when — including support access to your account.
Straight answers
What this product does and does not do
Unusual to publish. More useful than discovering a gap in week two of a trial.
| Capability | Status | Detail |
|---|---|---|
| Auto-scored objective questions | Yes | Single and multiple choice, fill in the blank, image-based and code-snippet MCQ are scored on submission. |
| Sandboxed code execution | Yes | 13 languages, run against your test cases with per-question time and memory limits. |
| SQL questions | Yes | Written as a coding question executed against SQLite and checked on the output your query produces. There is no schema designer or result-set diff view. |
| Manual marking for long-form answers | Yes | Comprehension, long answer and case study responses go to a review queue for a human to score. |
| Automatic scoring of long-form answers | No | Prose is not auto-scored. A reviewer marks it against your rubric, and the result records who scored it. |
| File upload or video-response questions | No | Not a supported question type. Candidates answer in the browser or in the code editor. |
| Assessment generated from a job description | No | Assessments are built by a person, from your question bank or a randomising pool. There is no generator. |
| Proctoring with recorded evidence | Yes | Webcam and screen snapshots, a timeline of events, and a weighted risk score linked back to the events behind it. |
| Plagiarism detection across the cohort | Yes | Code submissions are compared by token sequence, syntax tree and structure, with matching regions shown side by side. |
| Candidate-facing white labelling | Yes | Your logo, colours and support address across the candidate experience and its emails, on paid plans. |
| Independent security certification | No | We hold no SOC 2, ISO 27001 or similar attestation, and we will say so plainly until we do. The controls we do operate are listed on the security page. |
Run your first technical round this week
Create an account, build an assessment and invite candidates. Full capabilities on trial, no card needed.
Have questions? Email sales@parikshafy.com