Technical assessment with live proctoring: a guide for hiring teams

A candidate submits a perfect coding test, but did they write the code? A technical assessment with live proctoring is a coding or technical skills test monitored as it happens, by automated checks, a person, or both.

You get the score plus a record of how the test was taken: who sat it, what was on screen, and whether anything broke the rules. Read on to see what live proctoring catches and how to keep it fair for candidates.

TL;DR:

• A technical assessment with live proctoring is a coding test monitored in real time by software, a person or both. You get the score plus a record of how it was taken.
• Unmonitored remote tests no longer prove authorship. Stack Overflow found 51% of professional developers use AI tools daily.
• Gartner found 6% of candidates admitted to interview fraud and predicts 1 in 4 candidate profiles could be fake by 2028.
• Webcam alone underdelivers. Pair browser checks like tab and copy-paste tracking with photo ID verification and good question design.
• Hybrid proctoring suits most coding tests: software flags events in real time and a person reviews them before any decision.
• Scale checks by stage. Use light checks for a 10 to 20 minute screen and full proctoring for the 45 to 90 minute coding test.
• Never reject on a flag alone. Tell candidates which checks are on, offer accommodations and ask shortlisted people to explain their own code.

What is a technical assessment with live proctoring?

A technical assessment with live proctoring is a skills test for a technical role, such as a coding task, a debugging exercise or a SQL test, that runs under real-time monitoring. The monitoring checks identity at the start, watches the screen, webcam and browser during the session, and leaves a timestamped record your team can review before deciding.

"Live" can mean two things, and it helps to be clear which one you are buying:

• Live automated checks: Software watches the session as it happens and acts on the spot: it warns the candidate, logs the event, or ends the test when a rule is broken. Nobody has to watch the video.
• A live person: An interviewer or proctor is present in real time, usually in a live coding interview, and can ask the candidate to explain a line of code or share their screen again.

Most hiring teams use both: automated live checks for the take-home or timed coding test, and a person in the room for the final technical round. For a general explainer on the exam-style version, see our guide to live proctoring.

Why are hiring teams adding live proctoring to coding tests?

Hiring teams are adding live proctoring to coding tests because an unmonitored remote test no longer shows who wrote the code. AI assistants can produce a working answer in seconds, and some applicants are not who they say they are.

A clean score now needs evidence behind it. Two shifts drive this.

First, AI help is everywhere in software work. According to the Stack Overflow Developer Survey 2025, 84% of respondents use or plan to use AI tools in their development process, and 51% of professional developers use them daily.

That is fine on the job, but it means a timed coding test with no rules about AI measures the tool, not the person. Second, candidate fraud is rising.

A Gartner survey of 3,000 job candidates, published in July 2025, found that 6% admitted to interview fraud, either posing as someone else or having someone pose as them. Gartner predicts that by 2028, 1 in 4 candidate profiles worldwide could be fake.

For remote engineering roles with access to code and customer data, a proxy hire is a security problem, not only a bad hire.

Key takeaway: Live proctoring does not make a weak test strong. It makes sure that the result you are reading belongs to the person you are about to interview.

How does a technical assessment with live proctoring work?

A technical assessment with live proctoring works in three stages: identity and setup checks before the clock starts, live monitoring while the candidate codes, and a review of the score next to the integrity record after submission. Each stage answers a different question, from "is this the right person?" to "can I trust this result?"

Before the test

The candidate agrees to the rules, passes a system check, and confirms identity, for example with a photo ID matched to their face on camera. Full-screen mode, webcam and screen sharing are switched on.

During the test

The platform tracks tab switches, time spent outside the test window, copy and paste, extra monitors, faces on camera and talking. Serious events trigger a warning, and repeated ones can end the session automatically.

After the test

The hiring team sees the code score and a proctoring report side by side: webcam snapshots, screen recording, question-level activity logs and any flags with their timestamps.

Before any decision

A person reviews flagged sessions. A flag is a reason to look closer or ask a follow-up question in the next round, not an automatic rejection.

Pro tip: Tell candidates exactly which checks are on before they start. Clear rules cut accidental violations, such as a candidate opening documentation they thought was allowed.

What does live proctoring catch in a coding test?

Live proctoring catches the most common ways candidates get outside help in a remote coding test: AI tools, a second person, copied answers and someone else sitting the test. Each method leaves a different signal, so a mix of checks works better than webcam footage alone.

With 51% of professional developers using AI tools every day, according to the Stack Overflow Developer Survey 2025, AI help is the first thing to plan for. HackerEarth's February 2025 review of cheating in online technical assessments names AI-generated answers as the dominant method, followed by proxy candidates and remote help, and warns that webcam-only proctoring underdelivers.

That matches what the table shows: the strongest setups pair browser checks with identity checks and good question design.

Live, AI or hybrid proctoring: which suits technical hiring?

Live, AI and hybrid proctoring suit different volumes and stakes. AI proctoring scales to every applicant at low effort, a live person reads context that software misses, and hybrid proctoring lets software flag events while people review them.

Most technical hiring teams end up with hybrid. Read our full comparison of AI vs human proctoring if you are deciding for a whole program rather than one role.

Which proctoring level fits each stage of technical hiring?

The right proctoring level rises with the stakes of each stage. Light checks keep early screening friendly for hundreds of applicants, a fully proctored coding test protects the main skills decision, and a live round with a person confirms that the candidate can explain the code they submitted.

Stage 1: Screening quiz (10 to 20 minutes)

Use full-screen mode, tab proctoring, copy and paste blocking, question randomization and an honesty agreement. Keep friction low, because most applicants drop out here for skill reasons, not cheating.

Stage 2: Coding test (45 to 90 minutes)

Add webcam snapshots, photo ID verification, AI assistance detection and mandatory screen sharing. This is the score you will act on, so this is where proctoring earns its keep.

Stage 3: Live technical round (45 to 60 minutes)

A person runs a live coding or pair-programming session and asks the candidate to walk through their stage 2 answer. Nothing replaces hearing someone explain why they wrote what they wrote.

Pro tip: Ask every shortlisted candidate to explain one part of their submitted code in the live round. It is the cheapest way to confirm authorship, and honest candidates usually enjoy it.

How do you set up a technical assessment with live proctoring in Testlify?

Setting up a technical assessment with live proctoring in Testlify takes four steps: pick or build the coding test, switch on the proctoring checks the role needs, set your violation tolerance, and invite candidates from your ATS. Proctoring is included, with a few extra checks paid per attempt.

Step 1: Choose the test

Start from the coding tests in the 3,500+ test library, or generate a custom assessment from a job description with the AI test builder. Coding questions run in 45+ programming languages, and multi-file projects run in a proctored VS Code environment for Node.js, Python and Java.

Step 2: Switch on proctoring

Testlify has 20+ proctoring checks, including full-screen enforcement, tab proctoring, mouse-out tracking, multi-monitor restriction, webcam snapshots, session recording, face detection, talking detection, AI assistance detection, copy and paste blocking, IP and location restriction, question randomization, mandatory screen share and a live environment check that uses the candidate's phone as a side camera. Photo ID verification with face match and live video proctoring with screen recording are available per attempt.

Step 3: Set your tolerance

Decide how many seconds outside the tab are allowed, how many violations trigger a warning, and when a session ends automatically. Add a custom honesty agreement and send private invites so links cannot be shared.

Step 4: Invite and review

Send invites from your ATS through 100+ integrations such as Greenhouse, Lever, Workday and SAP SuccessFactors. Each report shows the score, per-skill breakdown, AI-written insights, timestamped logs and the proctoring record, so reviewers see the result and the evidence on one screen.

For the final round, Testlify supports live coding and pair-programming interviews, and Interview-as-a-Service gives you vetted domain-expert interviewers if your engineers are short on time. See every check in detail on the anti-cheating and proctoring features page.

How do you keep proctored technical assessments fair to candidates?

Proctored technical assessments stay fair when candidates know the rules in advance, can ask for adjustments, and are never rejected by a flag alone. Proctoring should protect honest candidates from being outscored by cheaters, not make every candidate feel like a suspect.

• Give notice and get consent: Tell candidates in the invite which checks are on (webcam, screen share, ID) and why. Testlify's honesty agreement records that they read the rules.
• Offer accommodations: The ADA.gov guidance on AI in hiring says employers should tell applicants what technology is used and how they will be evaluated, and provide reasonable accommodations, such as extra time or an alternative format, so a tool does not screen out a qualified person with a disability.
• Review every flag by hand: Talking detection can fire when a candidate reads a question aloud, and a glance away can be a dog, not a helper. A reviewer should watch the flagged moment before it counts.
• Collect only what you need: Turn on the checks that match the stakes of the stage, keep recordings only as long as your policy allows, and choose a vendor with clear data controls. Testlify holds ISO 27001 and SOC 2 Type II and supports GDPR, with regional data storage.
• Allow for bad connections: A dropped connection is not a violation. Set tolerance thresholds so one network blip does not end a session.

For more on keeping later rounds honest, read how to prevent cheating in online interviews.

What should you check before choosing a proctoring setup for coding tests?

Before choosing a proctoring setup for coding tests, check five things. A tool that misses one of them usually pushes the work back onto recruiters. Use this list when you compare options:

1. A real coding environment: Code should run in the browser with syntax highlighting, test cases and the languages your team uses. A text box is not a coding test.
2. Layered checks you can switch on per stage: Browser checks, webcam, screen share, ID and AI assistance detection, each with its own tolerance, so a screening quiz and a final test can differ.
3. Identity verification: Photo ID with face match matters most for remote roles with access to production systems.
4. Evidence next to the score: Reviewers need timestamps, snapshots and recordings tied to the question the candidate was on, not a single "integrity" number.
5. ATS and workflow fit: Results should sync back to your ATS and trigger the next stage without manual exports.

If you are comparing tools for a whole engineering hiring program, our page on the technical assessment platform for IT teams covers the wider requirements, and our guide to technical screening explains where proctored tests sit in the funnel.

Start running proctored technical assessments

A technical assessment with live proctoring gives your team two things a plain coding test cannot: confidence that the right person took it, and evidence to back every shortlist decision. Start light at screening, proctor the main coding test fully, and finish with a live round where candidates explain their own code.

Testlify includes 20+ proctoring checks, runs coding tests in 45+ programming languages and connects to 100+ ATS and HR tools. Book a demo to see a proctored coding test from the candidate's side, or start a 7-day free trial with no credit card.

Frequently asked questions (FAQs)