Skip to main content

AI screening with human decisions, built for Canadian hiring.

For Canadian staffing agencies and high-volume employers, TalentInitiator takes on the first stretch of hiring: sourcing through channels candidates agreed to, screening conversations, asynchronous interviews and one place to review it all. Your recruiters make every decision and record why.

It works in English and French, keeps candidate records in Canada and is priced by users and level of access, in Canadian dollars.

The recruiter’s Today screen: decisions waiting for a reason, outcome notices coming due and an interview ready to review. Illustration of the product, with fictional people and companies.

From job order to shortlist

Four parts that work as one workflow. Use them together, or start with the one that saves your team the most time.
  1. Sourcing

    Publish to your careers page and job boards, forward applications by email, import résumés or let candidates bring their own LinkedIn export. Every record keeps its source and legal basis.

    How sourcing works 
  2. Candidate conversations

    Outreach drafted in each candidate’s language and approved by a recruiter, then a short screening chat that discloses the AI up front and offers a person at every turn.

    How conversations work 
  3. Virtual interviews

    Candidates answer at their own pace in writing. Voice and video answers come with release 1.0 (target Q4 2027); the demo shows them with sample transcripts. Answers are evaluated from the transcript with quotes, never from faces or tone of voice.

    How interviews work 
  4. Candidate data hub

    One record per candidate with sources, consents, conversations, interviews, scores and decisions, plus an audit trail and retention rules you can show a client or an auditor.

    How the data hub works 

AI recommends. A recruiter decides, and says why.

Rankings, summaries and scores are labelled as recommendations. Moving a candidate forward, holding or turning them down always takes a person.
  • Every advance, hold or rejection needs a reason code and the recruiter’s own words.
  • A rejection whose reason only points at the AI recommendation, or repeats its wording, is refused by the system.
  • Scores come with the exact quotes they rely on, from the résumé, the chat or the interview transcript.
  • Candidates can read why a decision was made and ask for another person to review it.
See the whole workflow 
The decision drawer: the AI recommendation and its evidence beside the recruiter’s decision and reason. Illustration of the product, with fictional people and companies.

Built for two kinds of hiring teams

The same product, set up for the way each team works.
  • Staffing agencies

    Turn a client’s job order into a compliant bilingual posting on your own careers page, triage the applicants, and send clients an anonymized shortlist through a private link. Keep the ATS you already run.

    For staffing agencies 
  • High-volume employers

    Handle hundreds of applicants per opening without losing anyone: knock-out questions, ranking with evidence, outcome notices sent on time, and Ontario and Quebec rules checked before anything is published.

    For high-volume employers 

What candidates can count on

Someone applying through TalentInitiator knows when AI is involved, and always has a way to reach a person.
  • They are told about AI before the first question, in plain language.
  • They can ask for a person at every step, from the chat to the interview.
  • They answer at their own pace on a phone, in writing, and by voice from release 1.0 (target Q4 2027).
  • They hear the outcome within 45 days of their last interview.
  • They can read why a decision was made and ask for a human review.
The screening chat on a phone: the disclosure comes first, and “Talk to a person” stays on screen. Illustration of the product, with fictional people and companies.

Canadian hiring rules the product applies

The product applies these rules where the work happens: in the posting composer, the outreach queue and the decision drawer, not in a policy binder.
Rule setWhere it appliesWhat the product doesIn force since
Ontario job postingsOntarioAI-use statement, a pay range no wider than $50,000, no Canadian-experience requirement, the existing-vacancy statement and an outcome within 45 days of the interview.January 1, 2026
Quebec Law 25QuebecFrench first, profiling off by default, an assessment before information leaves Quebec, privacy impact assessment templates and a human review on request.September 22, 2023
Anti-spam law (CASL)Canada, federalA recorded consent basis before any outreach, sender identification in every message, and unsubscribe requests honoured at once.July 1, 2014
Federal privacy law (PIPEDA)Canada, federalConsent at first contact, access and correction for candidates, and every sub-processor disclosed with its location.
Accessibility (AODA)OntarioAccommodation offered in postings and interviews, and screens built to WCAG 2.1 AA.January 1, 2021

This describes how the product supports these rules. It is not legal advice.

Read the compliance overview 

Candidate records stay in Canada

Records, documents and interview recordings are stored in Canada, in Montréal. Recorded interviews will be transcribed in Canada; they are offered once that service is connected.

We use AI services in the United States, under contract, to read résumés, draft outreach messages, run the pre-screening chat and evaluate answers and transcripts. A résumé is read as submitted, contact details included; profiles, answers and transcripts are evaluated with names and contact details removed. None of it is used to train AI models.

Email and text messages are delivered by the providers listed on our sub-processor page.

Pilots run on the demo today, with fictional people and companies. Real candidate data comes with release 0.2 (target: Q1 2027), once sign-in for customer accounts, key management in Canada and delivery reports from the messaging providers are live.

Where backups are keptConfirmation pending

See every sub-processor and where it runs 

Run a hiring week with demo data

Sign in as Dana, a recruiter at a fictional Toronto agency, and take a personal support worker job from intake to shortlist. There is nothing to install, every person and company in it is fictional, and the AI evaluation runs offline.