Ribbon
Ribbon interviews your applicants by phone or video and scores each one. A recruiter reads the interview and decides. Ribbon publishes its price. $499 a month covers two recruiters and 100 interviews. Its terms let it train its AI on your interview recordings. Get an opt-out written into your order form.
Ribbon publishes its price, and a recruiter decides every hire
Ribbon interviews and scores every candidate. A recruiter decides who moves forward. Its terms let it train its AI on your recordings.
You have more applicants than your recruiters can phone. You want to start now, not run a long buying process. You are fine with a recruiter opening every interview before a candidate moves forward or gets turned away.
You need to know the candidate really is who they say they are. You need a written promise that the service stays up. Your legal team will not accept a contract that lets the recordings train Ribbon’s own AI.
$499 a month, billed yearly, covers two recruiters, two open roles and 100 interviews a month. Go over that and it charges $4.00 an interview. Paying yearly saves 20%.
A recruiter votes on every candidate before anything happens
The applicant gets a message with a link inside it. The link works once and expires the moment they start. All they need is a phone or a computer with a modern browser. They take the interview whenever suits them, not at a time your recruiter booked. Ribbon holds a real conversation instead of reading out a list of questions. It follows up when an answer needs more detail. It can do this in 10 or more languages. A screening interview starts with five questions by default.
Your team can keep the default questions, or write your own in any order. Every interview is then scored out of five. Ribbon’s default score covers three things. It checks how clearly the candidate speaks, how interested they seem, and the background they bring. You can add your own scoring points for the role. Write them yourself, or let Ribbon draft them from your job description. You can change what you score for halfway through a role. Ribbon then rescores every finished interview against the new points.
Every interview produces a recording, a transcript, a summary and the scores. One of your recruiters watches or listens to it. They read the transcript and the summary. They check anything the cheating checks flagged. They compare candidates side by side. Then they cast a hire or no-hire vote. That vote is the only thing that moves a candidate forward. Ribbon has no switch that advances or rejects a candidate for you. HeyMilo, Apriora and Humanly do have that switch, once someone on your side turns it on.
The cheating checks, which Ribbon calls integrity checks, listen to the room and watch the screen. They can catch a second voice or whispering, and a computer voice reading answers aloud. They can catch answers that come back too fast, or after a pause long enough to be reading. They can catch a script read word for word, a sudden change in vocabulary, or eyes leaving the screen. Ribbon does not try to read a candidate’s mood from their face or the tone of their voice. Reading emotion this way at work has been banned in the EU since 2 February 2025.
What reaches your hiring system depends on which one you use. Ribbon says it connects to more than 60 tracking systems. It names twelve. They are Greenhouse, Lever, Workday, Workable, Ashby, BambooHR, Bullhorn, SAP SuccessFactors, UKG, SmartRecruiters, Personio and Teamtailor. It can pull your open roles in automatically. When a candidate finishes, it sends back the interview status and the overall score. That happens only if your system has a place to hold a score. It also sends a link to the interview itself, so anyone who wants to hear it clicks back into Ribbon.
Our guide to AI interview agents covers how the rest of this market handles the decision.
Its price, flag guidance and a yearly outside bias audit come standard
You can work out the bill before any sales call
Ribbon publishes its plans. Each one lists the price, how many interviews it covers, and what you pay once you go over. You can check your own hiring numbers against a real price. Every paid plan starts with a seven-day free trial. You can cancel it at any time.
Ribbon tells recruiters what a flag does not mean
Ribbon tells recruiters to watch a flagged moment before they decide anything. It names reasons a flag can go off by accident. They include a poor connection, a candidate pausing to think, a candidate speaking a second language, or a noisy room. It asks recruiters to weigh a flag alongside the scores, the answers and their own impression. Share this guidance with whoever writes your hiring policy.
The bias audit is yearly and run by an outside firm
Ribbon’s data processing agreement promises an outside bias audit every year, run by Holistic AI. The last one found no unfair impact on any group. An outside firm has also checked its security design, in a report called SOC 2 Type I. A longer audit of how its security runs over time, called Type II, is under way. The audit and the report help you meet New York City’s Local Law 144. You still have to hold the audit, publish it and answer for it. That duty falls on the employer, not the vendor.
Candidates can interview any time, under your brand
It carries your own branding on every paid plan. Ribbon reports more than 500,000 interviews completed across all its customers.
Ribbon can train on your recordings and end service without notice
Ribbon AI Inc. is a young company based in Toronto. It announced $8m of funding in March 2025, led by Radical Ventures. Social Leverage, Cadenza Ventures and GD1 also took part. That is the only funding round it publishes. No acquisition and no layoffs have been reported. No case on the public court record names Ribbon. You are signing with a small supplier. Read the terms yourself before you sign.
Your interview recordings can train Ribbon’s AI
Ribbon’s terms cover the recordings, transcripts and answers you submit. They give Ribbon a worldwide right to use, copy, change and share that content. That right lets Ribbon train its AI models on it. Your applicants only agreed to an interview.
The contract can end your access at any time, without notice
Ribbon does not promise the service will always work, stay secure or run error-free. Its terms let it suspend or end your access at any time. It does not have to give a reason or warn you first. If something goes wrong, the most it will pay back is what you paid in the past twelve months. Any dispute goes to the courts of Toronto, under the law of Ontario. A named contact and a service promise come only on its custom-quoted plan.
It checks email, connection and location, but not identity
Your team can run a check on any candidate. It flags anything odd about their email address, their internet connection or their location. You choose when to run it. Only your team sees the result. None of this proves who the candidate actually is. Ribbon asks for no identity document and matches no face against one. It has no check for AI-faked video or voice.
What a candidate is offered instead is left to you
California and New York require you to let candidates opt out. Both laws also require you to “provide alternative accommodations”. Ribbon offers no right to a human interviewer instead, anywhere. If your hiring policy promises a real person on request, or an adjustment for a disability, you provide it yourself.
It suits teams with more applicants than recruiters can phone
Stepful put 2,500 applicants through instead of 40 by hand
Stepful screens admissions for health training courses. It has more than 25,000 active learners a month. Before Ribbon, its team could do 40 interviews by hand. With Ribbon, it put 2,500 applications through the tool instead.
A one-person HR team saved over 100 hours of screening
For more than a year, Cathy ran HR and talent by herself at Opencare, a healthcare marketplace in Toronto. She handled sourcing, screening and hiring for every open role. She says Ribbon saved her “over 100 hours of screening time” across more than 200 interviews.
Knix staffed three new stores in three weeks
Knix is an intimate apparel retailer opening stores across North America. It cut the time to screen one candidate from twenty minutes to five. It staffed full sales-associate teams for three new store openings in three weeks. LAZ Parking used to run a whole phone-screen team. It now reports the wait between an application and an interview fell from about 48 hours to under five minutes.
First Brands Group cut the time to fill technical roles by 65%
First Brands Group makes automotive parts under brands including FRAM, Autolite and Raybestos. It evaluated more than 800 engineering candidates across those brands. Recruitment costs fell 45%. 90% of its hiring managers were happy with the candidates they saw. Katalyze AI, a biopharma manufacturing startup, cut the time to hire for specialised roles from 40 days to 16.
Published plans run $499 to $1,999 a month, billed yearly
The next plan up costs $999 a month, for five recruiters and 400 interviews. The largest published plan costs $1,999 a month, for ten recruiters and 1,000 interviews. The charge for going over your plan drops too. It runs from $4.00 an interview on the entry plan, to $3.00, then $2.50, on the bigger plans.
These prices assume yearly billing. Price the month you run most often, not your peak month. If about 400 interviews is your usual month, the entry plan adds $1,200 to your $499 a month. The $999 plan that covers 400 interviews costs less in that case. If 400 interviews is a once-a-year spike, the entry plan plus that month’s extras costs less instead.
A custom-quoted plan sits above these three. It adds staff login with their existing company account, separate teams inside one account, and security reviews. It is invoiced, not priced on the website.
There is no minimum volume. The smallest published plan covers two recruiters and two open roles at a time. No named customer has a published headcount. Ribbon calls Opencare “a small but ambitious organization”. Its own co-founder calls Katalyze AI “a small hiring team”.
By default, Ribbon keeps recordings and transcripts for up to 24 months. You or the candidate can ask for them to be deleted sooner. Your account data lasts as long as the contract does. When the contract ends, Ribbon’s data processing agreement commits it to delete or return all personal data, whichever you choose. You can also cut off access to one interview’s recordings on demand.
You could also build a screen, an interview and a debrief on your own tracking system. We have mapped what that looks like. See it as an agent
Eight questions for the demo
- Is the consent message switched on for our account from day one? Who on our side can switch it off? Good answer: Only one named admin can change it, and every change is logged.
- Will you remove the clause that lets you train your AI on our recordings? Good answer: Yes, in the order form, and for recordings you already hold.
- What notice period and uptime figure will you put in our contract, without us buying the custom plan? Good answer: A written notice period and uptime figure on our own plan.
- Will you give us the full bias audit report and the figures for roles like ours? Good answer: The full report, with figures we can publish for Local Law 144.
- Where do your outside providers, LiveKit, OpenAI and Kombo, process and store our candidates’ data? Good answer: A named country for each one, and a line we can give a candidate who asks.
- Which of our systems can you send a score into, not just read jobs from? Good answer: A named list of systems, and what appears on a candidate’s record.
- What does a candidate get instead if they cannot take an AI interview? Good answer: A named alternative that does not count against the candidate.
- How do our recordings, transcripts and scores come out in bulk when the contract ends? Good answer: A stated file format and a time window, written into the contract.
HeyMilo, Apriora and Humanly can act alone. Ribbon cannot.
If Ribbon comes off your shortlist, HeyMilo, Apriora and Humanly do the same job. All three are sold for the same high-volume hiring. All three quote against your headcount or your applicant numbers, rather than publishing a figure. Employment Hero and Metaview Screening sit inside products some companies already pay for. Buying nothing is a sixth route.
Look at HeyMilo, Apriora and Humanly if you want the interview to act on its own result. Each can be set to move a candidate on by itself, without a recruiter opening the interview first. HeyMilo can also reject a candidate the same way. Someone on your side owns that setting. Apriora also adds fraud scoring, interview monitoring and an identity check against a government document. Ribbon has no such identity check.
Employment Hero also publishes a price. A$199 a month covers 50 interviews. Buying it does not mean moving your HR or payroll systems across as well. Metaview Screening charges per finished interview. It starts every account with 25 free interviews, so a trial costs close to nothing.
Buying nothing this year is the last option. At about forty screens a month, most of Ribbon’s 100 interviews would go unused. From 2 December 2027, hiring systems like this one count as high risk under EU rules, with duties attached. The UK’s new rules on automated decisions came into force on 5 February 2026. A person must now be part of any such decision. Ribbon keeps a person in the decision, as long as your recruiters really open the interviews. Our market guide covers building a narrower version yourself.
https://www.ribbon.ai/ · Ribbon homepage, including its questions and answers · read 19 Sep 2026
https://www.ribbon.ai/ai-interviewer · Ribbon AI Interviewer · read 19 Sep 2026
https://www.ribbon.ai/integrity-monitor · Ribbon integrity and cheat detection · read 19 Sep 2026
https://www.ribbon.ai/metrics · Ribbon reporting · read 19 Sep 2026
https://www.ribbon.ai/integrations · Ribbon integrations, the named systems · read 19 Sep 2026
https://www.ribbon.ai/about · Ribbon about us · read 19 Sep 2026
https://www.ribbon.ai/regulations · Ribbon on California, New York and EU rules · read 19 Sep 2026
https://www.ribbon.ai/terms-of-service · Ribbon terms of service · read 19 Sep 2026
https://www.ribbon.ai/privacy-policy · Ribbon privacy policy · read 19 Sep 2026
https://www.ribbon.ai/sub-processor-list · Ribbon data processing agreement and the outside suppliers it names · read 19 Sep 2026
https://www.ribbon.ai/bias-audit · Ribbon bias audit page · read 19 Sep 2026
https://www.ribbon.ai/changelog · Ribbon changelog, including teaching it how to score · read 19 Sep 2026
https://www.ribbon.ai/releases/ribbon-verification-signals · Ribbon release note, the candidate check · read 19 Sep 2026
https://www.ribbon.ai/ribbon-raises-8m-to-fix-hiring · Ribbon funding announcement, 31 March 2025 · read 19 Sep 2026
https://www.ribbon.ai/blog/bloomberg-article · Bloomberg, “Job Interviews Enter a Strange New World With AI That Talks Back”, 17 June 2025, reposted on Ribbon’s blog · read 19 Sep 2026
https://www.osler.com/en/about-us/representative-work/ribbon-ai/ · Osler, Hoskin & Harcourt LLP, its own record of the financing · read 19 Sep 2026
https://www.ribbon.ai/blog/knix-case-study · Knix case study · read 19 Sep 2026
https://www.ribbon.ai/blog/first-brands-group-case-study-streamlining-technical-hiring-across-multiple-brands-with-ribbon-ai · First Brands Group case study · read 19 Sep 2026
https://www.ribbon.ai/blog/opencare-case-study · Opencare case study · read 19 Sep 2026
https://www.ribbon.ai/blog/stepful-case-study · Stepful case study · read 19 Sep 2026
https://www.ribbon.ai/blog/helping-katalyze-ai-hire-their-founding-team · Katalyze AI case study · read 19 Sep 2026
https://www.ribbon.ai/blog/how-ribbon-sped-up-thrives-recruitment-process-by-80 · Thrive case study · read 19 Sep 2026
https://docs.ribbon.ai/knowledge-base/overview.md · Ribbon for recruiters, what a recruiter does with an interview · read 19 Sep 2026
https://docs.ribbon.ai/knowledge-base/interview-flows.md · Ribbon interview set-up, questions, languages and links · read 19 Sep 2026
https://docs.ribbon.ai/knowledge-base/custom-scoring.md · Ribbon scoring, the default criteria and your own · read 19 Sep 2026
https://docs.ribbon.ai/knowledge-base/interview-settings.md · Ribbon interview settings, including the consent message · read 19 Sep 2026
https://docs.ribbon.ai/knowledge-base/integrity-monitoring.md · Ribbon integrity checks and its warnings about them · read 19 Sep 2026
https://docs.ribbon.ai/knowledge-base/ats-integrations.md · Ribbon on what is sent to your own system · read 19 Sep 2026
https://docs.ribbon.ai/docs/supported-languages.md · Ribbon developer documentation, the language list · read 19 Sep 2026
https://api.usaspending.gov/api/v2/search/spending_by_award/ · USAspending.gov award search, no award naming Ribbon · read 19 Sep 2026
https://www.courtlistener.com/api/rest/v4/search/ · CourtListener search, no case naming Ribbon · read 19 Sep 2026
https://www.ribbon.ai/pricing · Ribbon pricing · read 19 Sep 2026
https://www.ribbon.ai/pricing · Ribbon pricing, what it charges once you go past a plan · read 19 Sep 2026
https://www.nyc.gov/site/dca/about/automated-employment-decision-tools.page · New York City Department of Consumer and Worker Protection, automated employment decision tools · read 19 Sep 2026
https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX:32024R1689 · EU AI Act, Regulation (EU) 2024/1689, Article 5 · read 19 Sep 2026
https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX:32026R1744 · Regulation (EU) 2026/1744, amending Article 113 of the AI Act · read 19 Sep 2026
https://web.archive.org/web/20260904114636/https://www.legislation.gov.uk/uksi/2026/82/regulation/2/made · S.I. 2026/82, commencing section 80 of the Data (Use and Access) Act 2025 · read 19 Sep 2026