Dayforce's AI Interviewer Hears Your Notes. Ditch the Cheat Sheet

You got an invite to an AI screening interview, and your first thought was to keep a page of talking points open next to the call. With Ribbon, the AI interviewer Dayforce has agreed to buy, that page is the most likely thing to hurt you. It won't hurt you because the AI sees it. It'll hurt you because the AI can hear you reading it.

What Dayforce is buying from Ribbon

Dayforce is buying a voice interviewer that writes its own questions and scoring rubric from the job description, scores you from the transcript, and runs separate integrity checks for "unusual audio, gaze, and response-timing patterns." The price and closing date haven't been disclosed, and the deal isn't complete.

Ribbon is based in Toronto and run by co-founder and CEO Arsham Ghahramani. Dayforce's release cites more than 1 million completed interviews. Ribbon's own homepage now claims 2M+ interviews completed and a 12-minute average time to screen. I'll use Ribbon's figure when talking about scale, since it's the newer one.

Dayforce's David Ossip said in the release that "high-volume hiring shouldn't mean choosing between speed and a strong candidate experience." That phrase, high-volume hiring, tells you who is likely to get these screens first: frontline and contact-center roles, where a recruiter can't phone every applicant.

Dayforce hasn't said how the integrity flags affect screening decisions. Nobody outside the company knows yet, so the useful thing is to understand how the two systems work and prepare for both.

The score and the flag are two separate systems

Your score comes from what you said. Your integrity flags come from how you said it. They're produced separately, and then they land on the same recruiter summary.

The score is built from the transcript, so it's a content judgment. A polished voice earns you nothing on its own, and specific answers earn you nearly all of it. Each score pulls exact quotes from your answers and marks positive and negative highlights with timestamps.

The flags are behavioral. A recruiter sees integrity monitoring flags with timestamps they can jump to, right next to your score. Ribbon tells employers that a flag is "not a definitive judgment." But it sells the same feature as fraud detection: it "automatically flags off-screen activity and unusual response patterns before a candidate ever reaches your team," and a contact-center page promises to catch "someone's ChatGPT answers."

The four integrity checks

Ribbon's integrity monitoring falls into four categories. Here's what each one watches for, and how a cheat sheet sets it off.

Ribbon admits these checks produce false positives. A poor internet connection, a normal thinking pause, non-native speech patterns or a noisy room can all trip a flag. So the goal isn't to look perfect. It's to avoid creating the specific pattern a cheat sheet creates.

Why notes trip the flags even on a voice-only call

Reading from notes produces a signature the system catches without a camera: a gap while you search the page, then answers that sound read aloud. In a voice interview, that pattern hits two of the four checks.

Ribbon works on any phone. On a voice-only screen there's no gaze to track, so a lot of candidates assume the integrity checks don't apply and keep their notes open. But Response Timing and Input Consistency both still run, and a cheat sheet is close to a textbook trigger for each:

• The search pause. You hear the question, scan for the right paragraph, and the line goes quiet longer than a normal thinking pause.
• The read-aloud switch. Once you find it, your speech gets smoother and more even than the way you talk when you're thinking. That's the "scripted speech" Input Consistency is built to catch.
• The mismatch. Your answers on unexpected follow-ups sound halting while your prepared answers sound recited. Because each flag carries a timestamp, a reviewer can jump straight to the moment you switched.

So the usual AI interview gaze tracking advice (look at the lens, put your notes behind the webcam) only fixes the video case. It does nothing for timing or delivery, and on a phone screen those are the only checks that matter.

The two-second turn problem

Ribbon's AI replies after a pause of 0.5 to 2 seconds, so a long silence while you think can be read as the end of your answer. That's how a normal thinking pause turns into a cut-off answer and possibly a timing flag.

What follows is my inference from the documentation, not something Ribbon has published: say you're thinking, out loud. "Let me think about the best example for a second" does two things. It keeps the turn from passing to the AI, and it gives anyone reviewing a timestamped pause a reason for it in the transcript.

A few more habits help with turn-taking:

• Start with a one-sentence answer, then expand. A turn that ends early still holds your main point.
• Finish with a clear closing line ("That's the example I'd point to") so the AI doesn't jump in mid-thought.
• If retries are allowed, use one when you were cut off. Ribbon lets employers allow retries up to a limit, or one attempt only.

Where a flag can cost you before a human looks

The biggest risk isn't a recruiter reading your flag and rejecting you. It's a flag or a low-ranked summary quietly pushing you down the queue so nobody opens it. All of this happens before a human ever watches the recording.

One analyst put it plainly: "a candidate can lose the opportunity earlier if a ranked summary is never opened or an integrity flag silently delays them." For context, Greenhouse found in its own survey that 51% of AI interview completers hadn't heard back. That's not Ribbon's number, and it doesn't measure flags. It does show how often finishing an AI screen ends in silence, and an unexplained flag gives a busy recruiter one more reason to skip you.

Who is likely to send Dayforce AI interviews: Refolk's index

In Refolk's index, the recruiters who tag Dayforce mostly work at frontline-heavy employers. Of the 7 recruiters and talent acquisition staff in the US and Canada who list Dayforce as a skill, 5 work at that kind of employer, including Tesla, U-Haul, Magna International and North York General Hospital.

These counts are small because few recruiters tag software on their profiles. Read them as direction, not market size. The direction lines up with Ossip's "high-volume hiring" pitch: if you're applying for shifts at a plant, a hospital or a rental counter, you're the likelier audience for a Ribbon screen.

The HireVue comparison matters too. In the US, 3.2 times as many recruiters in the index tag HireVue as tag Dayforce, which suggests many candidates have done a HireVue screen and will assume a Ribbon one works the same way. On opting out, it doesn't. HireVue's AI Interviewer page says "Candidates can opt out" and lists non-AI paths. Dayforce hasn't published an equivalent for Ribbon.

How to pass an AI voice interview without a script

Prepare stories, not sentences. The rubric is generated from the job description and scored from your transcript, so the best preparation is matching your real experience to the posting and practicing it out loud until you can tell it without reading.

Because about 90% of customers keep the defaults, these AI screening interview tips should apply across most Ribbon invites:

1. Read the invite for the settings. Check whether retries are allowed and whether there's a follow-up "validation level." Those are the settings most likely to differ between employers.
2. Pull the rubric out of the posting yourself. Ribbon builds its questions from the job description, so each requirement in the posting is a likely question. Write down one real example for each.
3. Turn each example into four beats, not a paragraph. Situation, what you did, the number, the result. You remember beats. A paragraph is something you end up reading.
4. Rehearse aloud, twice, without the page. Record yourself on your phone. If you sound like you're reading, you memorized wording instead of the story.
5. Put the notes away before you dial in. Look over your beats right before the call, then close the tab. A cheat sheet you don't look at can't create a pause.
6. Fix the environment. Use a quiet room, a wired or strong connection, and no one else talking. Audio Environment flags anyone else's voice, and Ribbon lists a noisy room and poor internet as known false-positive triggers.
7. On video, look at the lens when you're thinking. Interview Focus watches gaze. Looking down at a desk reads the same as looking at a second monitor.

Step two is the tedious one: going through the posting requirement by requirement and finding the evidence in your own history. It's also the part Refolk does for you. Paste the posting and you get your resume rewritten around it, plus a fit score showing which requirements your history covers well and which it barely touches. The weak ones are the questions to rehearse hardest, because those are where a vague answer will show up as a negative highlight in the transcript.

How to ask for an accommodation before you start

If a disability affects your eye contact, speech pace, pauses or use of assistive software, ask for an accommodation before you begin the interview. The ADA still applies, the request doesn't need legal wording, and asking first beats explaining a timestamped flag afterward.

In January 2025, the EEOC deleted the page explaining how AI hiring tools can illegally screen out disabled applicants. The law didn't change: Title I of the ADA still bans AI tools from screening people out because of a disability. The American Foundation for the Blind's June 2026 report found participants describing assessments that interfered with their accessibility software, and the four Ribbon checks (gaze, timing, scripted speech, audio) are exactly the ones that can misread a screen reader user, someone who stutters, or someone who needs longer to process a question.

How to ask:

• Reply to the invite, in writing if you can. A request can be spoken or written. Writing gives you a record.
• Use plain language. For example: "I have a disability that affects my response timing and eye contact. Before I complete this interview, I'm requesting an accommodation, such as a live interview with a person or having integrity flags excluded from my evaluation."
• Send it before you open the interview link. Afterward, you'd be arguing against a flag that comes with a timestamp.

Check your local law too:

• Illinois: The AI Video Interview Act requires employers to tell you AI may analyze your interview, get your consent, and delete recordings within 30 days if you ask. Because Ribbon watches gaze, the video provisions matter for any Illinois job.
• New York City: Local Law 144 requires notice at least 10 business days before an AI tool is used.

Whatever you send, keep the reply.

FAQ

Will looking away from the camera get me rejected in a Dayforce AI interview?

Nobody outside Dayforce knows. Dayforce hasn't said how integrity flags affect screening decisions, and Ribbon tells employers a flag is "not a definitive judgment." Looking away can trigger the Interview Focus check on video, but it's the smaller risk. Long search pauses and read-aloud delivery trip the timing and scripted-speech checks, and those run on voice-only calls too.

What if the AI cuts me off mid-answer?

Ribbon's AI replies after a pause of 0.5 to 2 seconds, so a long thinking silence can end your turn. Say "let me think for a second" out loud to hold the turn. If the employer allows retries, use one on any answer that got cut short.

Can I opt out of a Ribbon AI interview?

HireVue publishes an opt-out and non-AI paths, but Dayforce hasn't published an equivalent for Ribbon. If you have a disability, ask for an accommodation in writing before you start, for example a live interview. Title I of the ADA still applies to AI hiring tools. In Illinois you also have consent and 30-day deletion rights, and in New York City you're owed notice 10 business days before the tool is used.