A side application of IWT

First-round interview, automated.

Same engine you'd use for a content assessment, pointed at a candidate instead of a student. The hiring manager writes the questions, AI conducts the interview, and you skim a transcript and a grade in 5 minutes — instead of running 30 phone screens to find your real shortlist.

Why we built this for hiring teams

Most first-round phone screens reveal "no" in the first five minutes. The next 25 minutes are a courtesy. Multiply by 30 candidates per role and the hiring manager's entire week is gone before the actual interviews even start.

IWT's interview engine — the one teachers use to replace papers with adaptive AI conversations — works just as well as a hiring screen. You design the rubric, the candidate takes a 15-minute interview on their own time, and you read a structured transcript with an AI grade attached. Easy "yes" pile gets a real human interview. Easy "no" pile never wastes anyone's time.

How it works

Step 1

Hiring manager builds the interview

Author 5–10 questions about the role. Build the rubric in plain English — what a strong answer sounds like, what a weak answer sounds like, what disqualifies. Same flow as setting up a teacher rubric.

Step 2

Candidate takes the interview

Async, on their schedule. Voice or text. ~15 minutes. Same proctoring layer as Content Assessment catches reading off-camera, AI-generated answers, or coaching mid-interview.

Step 3

Hiring manager reviews

Transcript, AI grade per question, holistic summary. Skim in 5 minutes. Easy yes, easy no, or watch the full video for the borderline cases.

What's different from generic AI hiring tools

→You write the questions and the rubric. The interview tests what your team actually cares about — not a vendor's idea of what makes a good candidate.
→Same engine as our education product. Compound-signals-only authenticity layer. Personal speaking style, eye-contact aversion, restarts, and pauses to think are explicitly NOT flagged. Reading-from-script is.
→No scoring black box. Every grade comes with reasoning and evidence quotes from the candidate's own words. Override anything. Add your note. Defensible if a candidate ever asks why.
→The video is yours, not ours. Configurable retention. No model training on candidate data.

Who this fits

Recruiters running volume

High-applicant roles where the bottleneck is screening hours, not the actual interview cycle.

Hiring managers on technical roles

When the question is "do you actually understand X?" and a 5-minute conversation answers it. Build a rubric once, run it on every candidate.

Talent ops on consistency & fairness

Same questions, same rubric, same grading process for every candidate. Audit trail by design. Easier to defend hiring decisions than ad-hoc phone screens.

Status

The engine is live and in production for the education side. We're onboarding hiring teams selectively while we build the recruiter dashboard and candidate-facing surfaces. If you want to be in that early group, get in touch.

Get in touch about hiring

For Hiring