Podium assessment platform
Invite candidates, run research-backed assessments, video interviews and 360s, and trust every result. Anti-cheating technology with AI agent detection, Confidence Scores and Facial Validation, all on an ISO 27001-certified platform.
Anti-cheating technology
Generative AI tools such as ChatGPT and Gemini, and now AI agents that can operate a browser on their own, have made it easier than ever to get outside help during an unsupervised online test. Tools that try to spot AI-written answers struggle to keep pace.
Podium takes a different approach. Rather than guessing which tool a candidate used, it looks at how the assessment was taken, so protection keeps working as new AI tools appear. Every flag comes with its evidence and is reviewed by a person.
Test forms assembled from calibrated item banks, so shared or harvested answers are far less useful.
A short integrity message before every test. Our research found it significantly reduced flagged sessions.
A Confidence Score built from how the test was taken: leaving the test window, response timing, copy attempts and signs of automation.
Optional Facial Validation that guards against proxy test-taking, outside help and second-device use.
Every supported session gets a score from 0 to 100 showing how confident the system is that the test was completed under standard, fair conditions, banded from Very High to Very Low and colour-coded in your participant tables. It is a behavioural index, not AI.
Supported assessments check for AI agents and automation software completing the test instead of the candidate. A result where an agent is detected is capped at Low confidence, one where it is possible at Medium, and both carry a robot icon so they stand out.
Optional webcam checks for remote, unproctored testing. Periodic snapshots, not video, are compared with a reference image, and AI image analysis flags no face, more than one face, a phone in view or a possible change of person. Review results per candidate or across a whole project.
Click-away events, exits from full-screen mode and copy attempts are recorded with timestamps, alongside time spent on each question compared with the norm.
Full-screen delivery, clipboard restriction and blocking of common keyboard shortcuts and developer tools make it harder to move information into or out of the test.
On supported assessments, replay a recording of the test page as the candidate saw it, with the questions and the candidate's answers hidden.
The Confidence tab explains what was found and when, including the innocent explanations, so you can follow up fairly before relying on a score. Export it to PDF for your records.
If a connection drops or a browser closes, candidates carry on where they left off, and each recovery is recorded with the rest of the evidence.
Integrity flags never trigger automated decisions. Every flagged session goes to a person, and a low score is a prompt to review, not proof of misconduct.
Our four-layer integrity framework was tested in research presented at the SIOP 2026 Annual Conference, covering a controlled experiment and more than 106,000 GCAT sessions in the field.
Security & compliance
Candidate data is sensitive. Podium is built and run under an ISO 27001:2022-certified information security management system, with the controls your IT, legal and procurement teams will ask about.
An independently certified information security management system covering how Podium is built, hosted and supported.
TLS 1.3 protects data in transit, and data is encrypted at rest.
Separate regional platforms, including Europe and the UK, Australia and New Zealand, and the UAE, keep your data in your region.
Protect administrator accounts with MFA and require it across your organisation, with configurable password-change intervals.
Administrators sign in through your organisation's identity provider using SAML 2.0 or OAuth 2.0.
Granular roles and permissions, including a restricted test administrator role for colleagues who only need to run assessments.
Timestamped activity logs for every participant and project, including email delivery records, exportable for audit.
Regular independent penetration testing and security assessments.
GDPR compliant. AI services run on zero-data-retention terms, AI summaries never include names or identifying details, and no AI scores, ranks or makes decisions about candidates.
Land on what matters today: pinned work, recent activity, alerts and credit usage, with search and an AI assistant always to hand.
Recent invitations and results, pinned items, alerts and credit usage on one screen you can arrange to suit you.
Search participants, invitations and results across the whole system, with shortcuts and filters for test, group, project and date range.
Ask questions about your data in plain English, build charts and prepare documents, with saved conversation history and optional spoken answers.
Raise a support request from the dashboard, attach screenshots or files, and follow the conversation without leaving Podium.
See invitations sent, results received and report bundles over your chosen date range, and export them to Excel.
Spot invitations that didn't arrive, correct the email address and re-send without leaving the dashboard.
A live credit balance with a date-ranged history of every credit added and spent.
Pin assessments, participants, groups and projects to your dashboard for one-click access.
Check credits, recent invitations and results, browse projects and groups, and add participants from your phone.
A participant workbench: search, filter, edit inline, open full records and act in bulk without leaving the page.
Inline editing and expandable records with tabs for results, confidence, Facial Validation, membership, personal details, timeline and files.
Sten scores and percentile equivalents shown against each participant, with scale descriptions inline.
Opt in to a one-click summary of a candidate's scores, clearly labelled as AI-generated and produced without names or identifying details.
Stack filters with date ranges, then invite, remind, export, group or archive the results in one go, including drag and drop.
Organise participants into groups with the same filtering, import, pinning and export tools.
Upload or paste participants from Excel or CSV.
Visual breakdowns of candidate bio data across each project to help you monitor fairness.
Authorise 20% additional time on timed assessments for individual candidates.
Store documents securely against a participant, and export details, scale scores, confidence data and bio data in full.
Run each campaign in its own workspace, with assessments, 360s, video interviews, shortlisting, cut-offs and analytics side by side, and the right way in for every candidate.
For supervised rooms such as computer labs, exam halls and assessment centres. Open the room from the venue's network, watch candidates arrive and progress live, and close it when the door shuts.
An open link with a guest list: only candidates whose email or your own identifier you supply can register. Describe the identifier's format in plain English and the AI builds the validation.
Short, memorable go.podium365.com links for Open and Restricted Link projects, one per language, easy to print on a poster or read out over the phone.
One shareable link for high-volume campaigns, with controls over who can register and when the link expires.
Send branded email invitations, schedule them in advance, set expiry dates and send reminders to everyone who hasn't finished.
Rank candidates by score with colour-coded badges, then request reports or move a shortlist into another project.
Set pass, reserve and fail bands per scale against sten, percentile or raw scores, tied into shortlisting.
Pick a test, scale and score type to see how your candidates are distributed, side by side.
Live progress for every candidate and a timestamped log of every project event.
Build, share and score one-way video interviews, and run fully tracked 360 and 180 feedback programmes.
Configure questions, timings and retakes for one-way video interviews, included alongside assessments in any project.
Score interviews in Podium or share secure links with external scorers, then export the results.
A dedicated 360 interface with rater management, progress tracking and reminders.
Import self-raters and raters from a spreadsheet and manage them in bulk.
Request the right reports in the right languages, watch them generate, and find any report ever produced in seconds.
Search and filter every report ever requested, open any bundle to see its participants, and download again from one place.
Preview what's in each report bundle and request several in one go.
Request reports in the languages your stakeholders read.
Watch reports generate in real time and see completion or errors as they happen.
Build ideal profiles for a role with AI-suggested sten scores for each scale, then fine-tune them yourself.
Set up batteries, emails and interviews with builder-grade tools, and keep data tidy with full archive and recovery.
Combine assessments into batteries you can reuse across projects.
A rich template builder for invitations and reminders, with defaults and language variants.
See the tests and batteries assigned to your account, with extra information on each.
See exactly who in your organisation you're sharing data with.
Archive participants, groups and projects by drag and drop, recover them with their data intact, and delete permanently only when you mean to.
Built-in guided tours plus written and narrated video tutorials for every key task.
Podium is designed so that it doesn't matter which AI tool a candidate tries. Supported assessments monitor how the test is taken (leaving the test window, response timing compared with the norm, copy attempts and signs that automation software or an AI agent is controlling the browser) and summarise it in a Confidence Score with the evidence. Optional Facial Validation guards against second-device use and proxy test-taking. In research presented at SIOP 2026, skill with AI tools did not help candidates evade detection.
A score from 0 to 100 showing how confident the system is that an assessment was completed under standard, fair conditions. It is a behavioural index, not AI, banded from Very High (90–100) to Very Low (0–39). A low score is a prompt to review the evidence and follow up, not proof of misconduct.
Yes. Supported assessments check for AI agents and automation software operating the browser instead of the candidate. Where an agent is detected the result is capped at Low confidence; where it is possible, at Medium. The Confidence tab explains the evidence in plain English, including innocent explanations.
No. Facial Validation is optional and takes periodic snapshots rather than recording video. AI image analysis compares each snapshot with a reference image and flags issues such as no face, more than one face or a phone in view, for a person to review.
No. Candidate scores come from established psychometric scoring against norm groups. No AI scores, ranks, shortlists or makes decisions about candidates, and no decision is made solely by automated means.
Podium runs under an ISO 27001:2022-certified information security management system. Data is encrypted in transit with TLS 1.3 and at rest, held on regional platforms, and protected by MFA, single sign-on, role-based access and regular penetration testing. Podium is GDPR compliant.
Yes. Network Link projects let anyone on the venue's network register without a PIN while the room is open, and a Live Session view shows every candidate arriving and progressing page by page.
Book a demo or sign in.
Assessment integrity research, product updates and practical advice on fair, secure hiring in the age of AI. Follow us on LinkedIn.