Candidates write, run and submit real code in the browser. Flowmingo scores it on every test case, including hidden ones they never see. AI drafts the problems from your job description. Unlimited technical tests, free.
Overall - AI Interview Score9.0Problem-Solving Skills9Marketing Campaign Execution9.5CRM Software Familiarity9
YouFlowmingoCould you please describe a time when you had to manage conflicting priorities between sales, marketing, and operations?Recording 0:42 · You can submit anytime
Overall - AI Interview Score9.0The candidate performed exceptionally well in the interview, providing detailed, specific, and quantifiable examples for nearly all requirements.Candidate's band50–60thRight around the middle for this interviewRecommendationsStrengthsGapsRecommend proceeding to the next stage of the hiring process.In the next interview, explore specific B2C sales experiences if relevant to the role’s scope.
Assessment Rubrics (9)Problem-Solving SkillsMust Have9The candidate provided a concrete example of mediating between sales and operations regarding delivery times and stock levels, identifying the core issue and proposing a solution that satisfied both internal stakeholders and the client.Vendor and Supplier Liaison8.5Marketing Campaign Execution9.5Communication and Negotiation Skills9CRM Software Familiarity9Digital Advertising Understanding9Sales Experience (B2B or B2C)8.5
AI InterviewCould you please describe a time when you had to manage conflicting priorities between sales, marketing, and operations?02:107/10How familiar are you with A/B testing, and how have you used it to optimize marketing campaigns or sales strategies?01:589/10What strategies do you use to stay up-to-date with the latest trends and best practices in sales, marketing, and operations?01:378/10TranscriptStrong point01:30So I stepped in and balanced both sides, I double checked with our operations to confirm that we can still deliver.
Overall - CV Evaluation Score8.8The candidate’s CV presents a very strong profile, demonstrating extensive and quantifiable experience across marketing, operations, and sales functions.RecommendationsStrengthsGapsProceed to interview to further explore specific negotiation scenarios and communication strategies.
The most secure multiple-choice test, proctored end to end. 20+ anti-cheating measures. Camera, mic and screen all monitored. Hands on device at all times.
Overall - Skill Test90%Correct18Incorrect2Unanswered0Time taken03:32 / 15:00CameraOn for the whole testTab switchesNonePasted textNone
TestDemo - Test Set (Marketing & Operation)Question 2 of 20RECTime19:33AI assistance in this exam is prohibited & carries serious consequencesAI assistance in this exam is prohibited & carries serious consequencesAI assistance in this exam is prohibited & carries serious consequencesAI assistance in this exam is prohibited & carries serious consequencesAI assistance in this exam is prohibited & carries serious consequencesAI assistance in this exam is prohibited & carries serious consequencesQA B2B SaaS startup spends $60,000 per month on marketing campaigns and $40,000 on sales team salaries. During the month, they acquire 50 new customers. What is the Customer Acquisition Cost (CAC) for this month?W1Answer 12Answer 2 · selected3Answer 34Answer 4
Task 1 of 301:12 / 05:00Walk us through a project you shipped. Share your screen.Recording your screenScreenBrowser tabCameraOnMicrophoneOnTime left03:48Stop and submit
Tasks3 of 3 recordedShowcase is not scored. The rating is yours.
Overall - Role-play Score7.0Listening and acknowledgement7Ownership9Alignment and close6
Role-playThanks for jumping on. I'll be direct: of the 120 leads marketing sent us last month, roughly 60% were students, consultants, or shops that couldn't pay for a seat.00:316/10That's the right number to track: cost per qualified lead, not cost per form fill.00:449/10TranscriptStrong point00:09So 122 leads, about 70 of them are people who can never buy, and your team spent the month finding that out.00:19I know that frustrated, and that’s a real cause, and that is one of the problems that you need to fix.
Overall - AI Interview Score9.0Problem-Solving Skills9Marketing Campaign Execution9.5CRM Software Familiarity9
YouFlowmingoCould you please describe a time when you had to manage conflicting priorities between sales, marketing, and operations?Recording 0:42 · You can submit anytime
Overall - AI Interview Score9.0The candidate performed exceptionally well in the interview, providing detailed, specific, and quantifiable examples for nearly all requirements.Candidate's band50–60thRight around the middle for this interviewRecommendationsStrengthsGapsRecommend proceeding to the next stage of the hiring process.In the next interview, explore specific B2C sales experiences if relevant to the role’s scope.
Assessment Rubrics (9)Problem-Solving SkillsMust Have9The candidate provided a concrete example of mediating between sales and operations regarding delivery times and stock levels, identifying the core issue and proposing a solution that satisfied both internal stakeholders and the client.Vendor and Supplier Liaison8.5Marketing Campaign Execution9.5Communication and Negotiation Skills9CRM Software Familiarity9Digital Advertising Understanding9Sales Experience (B2B or B2C)8.5
AI InterviewCould you please describe a time when you had to manage conflicting priorities between sales, marketing, and operations?02:107/10How familiar are you with A/B testing, and how have you used it to optimize marketing campaigns or sales strategies?01:589/10What strategies do you use to stay up-to-date with the latest trends and best practices in sales, marketing, and operations?01:378/10TranscriptStrong point01:30So I stepped in and balanced both sides, I double checked with our operations to confirm that we can still deliver.
Overall - CV Evaluation Score8.8The candidate’s CV presents a very strong profile, demonstrating extensive and quantifiable experience across marketing, operations, and sales functions.RecommendationsStrengthsGapsProceed to interview to further explore specific negotiation scenarios and communication strategies.
Overall - Skill Test90%Correct18Incorrect2Unanswered0Time taken03:32 / 15:00CameraOn for the whole testTab switchesNonePasted textNone
TestDemo - Test Set (Marketing & Operation)Question 2 of 20RECTime19:33AI assistance in this exam is prohibited & carries serious consequencesAI assistance in this exam is prohibited & carries serious consequencesAI assistance in this exam is prohibited & carries serious consequencesAI assistance in this exam is prohibited & carries serious consequencesAI assistance in this exam is prohibited & carries serious consequencesAI assistance in this exam is prohibited & carries serious consequencesQA B2B SaaS startup spends $60,000 per month on marketing campaigns and $40,000 on sales team salaries. During the month, they acquire 50 new customers. What is the Customer Acquisition Cost (CAC) for this month?W1Answer 12Answer 2 · selected3Answer 34Answer 4
Task 1 of 301:12 / 05:00Walk us through a project you shipped. Share your screen.Recording your screenScreenBrowser tabCameraOnMicrophoneOnTime left03:48Stop and submit
Tasks3 of 3 recordedShowcase is not scored. The rating is yours.
Overall - Role-play Score7.0Listening and acknowledgement7Ownership9Alignment and close6
Role-playThanks for jumping on. I'll be direct: of the 120 leads marketing sent us last month, roughly 60% were students, consultants, or shops that couldn't pay for a seat.00:316/10That's the right number to track: cost per qualified lead, not cost per form fill.00:449/10TranscriptStrong point00:09So 122 leads, about 70 of them are people who can never buy, and your team spent the month finding that out.00:19I know that frustrated, and that’s a real cause, and that is one of the problems that you need to fix.
AI writes the problems from your job description and checks each one before you see it. Candidates solve them in a code editor in the browser. Their code runs against test cases, and you read the score and the code behind it.
Sandbox checkThe AI's own solution passes. The problem is kept.
Describe the role
Paste a job description, upload one as a PDF or Word file, or list the skills you want to test. Pick a difficulty, from easy to very hard, and how many problems you want, from 1 to 6. The default is two, and one of them is a debugging problem unless you turn that off.
Check the problems
For each problem the AI writes the statement, example cases, hidden cases and starter code, then runs its own solution against the cases in a code sandbox. A problem that fails the check is dropped, so a test can come back with fewer problems than you asked for, and you can ask for a few more. This takes a few minutes.
You can edit the statement and the cases before you send the test; the builder re-checks your changes in the sandbox, and a problem that fails the check cannot be published.
Send it and read the results
Add the test to a hiring project, then invite candidates by email, upload a CSV of up to 100 names at a time, or share one link. The score appears on each candidate's row on your hiring board beside their other assessments, and you can filter the board by score range.
What candidates see
A code editor in the browser
Candidates write in a language you allow, starting from starter code. They install nothing.
function solve(input) {
// starter code
}
Run code, run tests and submit
Run code tries the code on the candidate's own input. Run tests checks it against the example cases. Submit scores it against every case, hidden ones included, and the candidate does not see the hidden results.
Run codeRun testsSubmit
One submit per problem
After a candidate submits a problem, its editor is locked, so the code you read is the code that was scored.
Submitted→Editor locked
A practice run first
Candidates are offered a practice run before the real test, when one is available, so they can try the editor first.
Practice run→The real test
Laptop or desktop only
Candidates need a laptop or desktop with a physical keyboard, in a current Chrome, Edge, Firefox or Safari. On a phone they see a message asking them to switch.
ChromeEdgeFirefoxSafari
Camera, microphone and screen
Candidates are asked to share their camera, microphone and whole screen after a consent step, and the session is recorded. A machine with no microphone can still take the test, without audio. A window or tab share is rejected when the browser identifies it; some browsers do not report what was shared.
CameraMicrophoneWhole screen
Your own questions, before and after
You can add questions that candidates answer before the test starts and after it ends. Both are optional.
Before the test…After the test
Scored by running the code
Example result
85%
17 of 20 test cases passed
Test cases passed per problem
Problem 1: 9 of 10
Problem 2: 8 of 10
How it is scored
The score comes from running the code, not from an AI's opinion of it. Each submission runs against every test case in an isolated sandbox with no internet access, and each problem earns points in proportion to the test cases it passes.
The example cases count as well as the hidden ones. Hidden cases are what stop code that only handles the examples from scoring full marks.
Only submitted code is scored. Code left in the editor when time runs out is saved but scores zero.
Your own systemsCreate tests, invite candidates and read scores through the API. A webhook tells you when an attempt has ended. API documentation
The score as a percentage and in points, with how many problems were solved, partly solved, failed or not attempted.
Test cases passed out of the total, and the minutes used.
For each problem, the graded code and the results for the test cases that ran, hidden ones included.
A score band comparing this attempt with eligible completed attempts on the same test; the band can change as more attempts are completed.
The recording of the camera, voice and screen, with integrity signals marked on a timeline. Recordings and timeline playback depend on your plan and on the recording being available.
Marked on the timelineTab switchPasteSecond window
Integrity signals
Signals for you to review, not verdicts.
Candidates are asked to share their camera, microphone and whole screen, and the session is recorded. Flowmingo logs the signals below. When the recording is available on your plan, they are marked on a timeline beside it, so you can jump to the moment and judge it.
What is logged
Switching tabs or apps, and how long the candidate was away
Pastes and copies, with their size but not their text
The same test open in a second window
Screenshot shortcuts
A browser side panel docked beside the test for a long time
Speech, which is detected but not transcribed, and a muted microphone
A screen share that stopped
Where the browser supports it, the candidate out of camera view or a second face in view
What it blocks, and what it does not
Pasting and switching tabs are not blocked.Candidates can do both, and you see that they did.
Capture is the one thing it enforces.A window or tab share is rejected when the browser identifies it. If a candidate stops sharing the screen or turns off the camera or microphone, a Monitoring required screen covers the test until it is back, and the clock keeps running.
It takes no time off the candidate's clock.Unlike the multiple-choice skills tests, no signal here costs the candidate time.
It is not proof of cheating.A signal points to a moment for a person to review.
AI assistance is off by default. Candidates are told in the instructions and again in the integrity check, and a watermark stays on screen during the test to deter AI assistants that read the screen. That deters; it does not enforce. You can switch AI assistance on for a test.
Free, with limits worth knowing
Included on every plan
Unlimited Technical Tests are included on every plan, including Free.
AI writes the problems from your job description and checks each one before you see it.
Hidden test cases, and the graded code for every problem in the report.
Problems and languages.When the AI writes a test, you ask for 1 to 6 problems. You choose the languages for each test from common ones such as Python, JavaScript, Java and C++. The test builder greys out any language that cannot run today.
Format.Each answer is one file, run in an isolated sandbox with no package installs.
Length.One technical test can run up to 60 minutes on Free, 90 on Starter and 120 on Growth and Scale.
Free sessions.On a free session candidates see Flowmingo's brand beside yours and may see optional offers. A sponsored technical test shows only your brand and no offers, and costs twice the rate of a sponsored interview.
Recordings.Whether you can play a recording, and how long it is kept, depends on your plan. A recording can also be unavailable for an attempt.
Beta.Technical Test is marked Beta in the app.
How it works
From job description to scored reports in 3 steps.
“The AI-powered interview process allows us to screen candidates consistently and efficiently, helping us identify strong talent much faster than traditional methods.”
Ryan AngTalent Acquisition, KFC
Agency
Send clients the interview, not just the CV.
“Flowmingo has significantly improved the quality of hires across our recruitment engagements.”
“The platform was easy to navigate, the instructions were clear, and I didn’t have to worry about complicated technical steps, allowing me to focus on answering the questions.”
“I found it very easy and simple to operate … The video interview feature was particularly valuable, as it was well-structured, useful, and organized, making the interview process smooth.”
“The AI interview platform itself was incredibly intuitive and user-friendly, effectively transforming what is typically a high-stress, nerve-wracking process into one that felt completely manageable and straightforward.”
“The platform loaded effortlessly across both desktop and mobile web browsers without requiring complex downloads or setups. Navigation was intuitive from start to finish.”
“What I appreciate most about Flowmingo is its candidate-centric user experience … it leverages intelligent automation to make the application process fast, transparent, and objective, replacing tedious manual forms with a seamless workflow.”
“Excellent user experience. It allows you to schedule and complete interviews at your convenience, completely eliminating conflicts with a busy schedule.”
“I really love how natural the AI interview process feels. It doesn’t feel like you’re just answering a rigid, robotic script; the system actually listens to what you say and adapts the follow-up questions dynamically based on your responses.”
“It provides fair questions related with the position and has a great feedback from the interviewer/avatar in communicating any doubt about the questions and topic.”
“I was quite pessimistic about the concept of an AI interview but upon completing it I realised that it was actually far more efficient for both candidate and employer.”
“I had a very positive experience overall. The platform is easy to use and well-designed, making it simple to navigate and complete tasks efficiently.”
Yes. Unlimited Technical Tests are included on every plan, including Free. On a free session candidates see Flowmingo's brand beside yours and may see optional offers. A sponsored technical test shows only your brand and no offers.
How does the AI write the coding problems?
You paste a job description, upload one as a PDF or Word file, or list the skills to test, and pick a difficulty from easy to very hard. The AI writes each problem's statement, example cases, hidden cases and starter code.
Does anyone check the problems the AI writes?
Yes. For each problem the AI runs its own solution against the cases in a code sandbox, and a problem that fails the check is dropped. A test can therefore come back with fewer problems than you asked for, and you can ask for a few more.
Can I edit the problems the AI writes?
Yes. You can edit each problem's statement and its example and hidden cases. The builder re-checks your changes in the sandbox, and a problem that fails the check cannot be published.
How many problems can a test have?
When the AI writes a test, you ask for 1 to 6 problems. The default is two, and one of them is a debugging problem unless you turn that off.
Which programming languages can candidates use?
You choose the languages for each test from common ones such as Python, JavaScript, Java and C++. The test builder greys out any language that cannot run today.
How is a coding test scored?
By running the candidate's code against the test cases in a sandbox. Each problem earns points in proportion to the cases it passes, and the example cases count as well as the hidden ones. Only submitted code is scored.
What are hidden test cases?
Cases the candidate never sees. Submit scores the code against every case, hidden ones included, and the candidate does not see the hidden results. They stop code that only handles the examples from scoring full marks.
Can candidates run their code before they submit?
Yes. Run code tries it on the candidate's own input, and Run tests checks it against the example cases. Submit scores it against every case. Each problem takes one submit, and its editor is locked afterwards, so the code you read is the code that was scored.
What happens when the time runs out?
Only submitted code is scored. Code left in the editor when time runs out is saved but scores zero.
Can candidates cheat with AI or by pasting code?
Pasting and switching tabs are not blocked. Pastes and copies are logged with their size but not their text, and tab switches, side panels, a second window and screenshot shortcuts are logged too. The session is recorded, so where your plan keeps the recording you can check the moment. AI assistance is off by default. Candidates are told in the instructions and again in the integrity check, and a watermark stays on screen to deter AI assistants that read the screen. That deters; it does not enforce. You can switch AI assistance on for a test.
Is the candidate recorded?
Candidates are asked to share their camera, microphone and whole screen after a consent step, and the session is recorded. Speech is detected but not transcribed, and camera view and a second face are checked where the browser supports it. If a candidate stops sharing the screen or turns off the camera or microphone, a Monitoring required screen covers the test until it is back, and the clock keeps running. Whether you can play a recording, and how long it is kept, depends on your plan, and a recording can be unavailable for an attempt.
Can candidates take it on a phone?
No. The test needs a laptop or desktop with a physical keyboard, in a current Chrome, Edge, Firefox or Safari. On a phone candidates see a message asking them to switch. They install nothing.
How do I send the test to candidates?
Add the test to a hiring project, then invite candidates by email, upload a CSV of up to 100 names at a time, or share one link.
What do I see in the results?
The score as a percentage and in points, how many problems were solved, partly solved, failed or not attempted, test cases passed out of the total, the minutes used, and for each problem the graded code. The score also appears on the candidate's row on your hiring board, beside their other assessments.
Can I add my own questions to the test?
Yes. You can add questions that candidates answer before the test starts and after it ends. Both are optional.
Can I create tests and read scores through the API?
Yes. You can create tests, invite candidates and read scores through the API, and a webhook tells you when an attempt has ended.
Is Flowmingo really free? What's the catch?
Yes. The full screening suite is free forever for companies, with unlimited AI Interviews, CV Evaluation, Tests and Showcase, a Full ATS and 60+ languages, and more assessment types on the way. Not a trial, not a teaser. No catch, no gotchas. If you ever want more power, longer assessments or your own branding, optional upgrades are there, but you never pay to screen a candidate.
How does Flowmingo make money?
Two optional ways. Candidates can choose to buy their own prep and career-feedback tools (never required, and our AI grader is blind to who paid). And companies who want more power or their own branding can upgrade to a paid plan. The core screening stays free, always.
Do I have to move off my current ATS?
No. Flowmingo includes a full ATS if you want one, but nothing requires you to use it. Paste the assessment link into the email your ATS already sends, invite candidates through the API, or have results posted back to your own systems by signed webhook. Your existing ATS keeps being the place you track candidates.
Why make it free for companies?
Because fair, high-quality hiring tools shouldn't be locked behind enterprise pricing. Our model lets us serve every team, from a solo HR manager to a global recruiting operation, without ever charging you to screen a candidate.
Do I need to talk to sales to get started?
No. Start in about 60 seconds on your own, and nobody will chase you. If you do want to talk, booking, WhatsApp and email are on this page — on the free plan too.
For candidates
Are you a candidate?
Get your CV in front of people who are hiring. It is free for candidates.