It notices hesitation before you do.
It adapts in real time — not after the fact.
It reports what actually happened.
Americans are neurodivergent — and most interview preparation tools weren't built with them in mind.
AdaptIQ closes that gap with real-time, adaptive AI coaching.
It notices hesitation before you do.
It adapts in real time — not after the fact.
It reports what actually happened.
Programs serving neurodivergent job seekers see this every cohort: strong candidates who freeze up, mask exhaustively, or get marked down for delivery instead of substance.
Perceive, process, and act — in real time, without adding a single manual step for your staff.
The browser locally reads facial landmarks, eye gaze, head pose, and voice during the session. Nothing raw ever leaves the candidate's device.
Lightweight behavioral signals — not video — are reasoned over by Claude to identify confusion, disengagement, or overwhelm as it happens.
The candidate gets a real-time, profile-matched intervention. After the session, staff get a structured report — not just a number.
ADHD, anxiety, and ASD-oriented profiles, each with distinct intervention styles and sensitivity thresholds.
Per-session breakdowns across eye contact, pacing, vocal confidence, and clarity, plus a summary staff can act on.
Candidates can run sessions on their own time — no scheduling a coach for every rep.
Interventions and answer feedback come from live model reasoning, not a static script.
"A candidate who used to freeze up in mock interviews walked into her first real one and actually finished her answers."
Give caseworkers a scalable way to prepare job seekers between one-on-one sessions, with data on who needs more support.
Support a full cohort of students without requiring a coach to sit in on every mock interview.
Extend the reach of neurodivergent employment programs without proportionally scaling headcount.
Offer candidates a genuinely accessible way to prepare, rather than a format that filters out neurodivergent talent by default.
Themed around social impact, inspired by Anthropic CEO Dario Amodei's essay Machines of Loving Grace.
Video and audio never leave the candidate's device unprocessed. Only abstracted behavioral signals — not footage — are sent to Claude for reasoning.
Face and voice processing happens client-side. Recordings are never uploaded or retained by default.
Only derived signals (gaze deviation, pace, tone) are sent for AI reasoning — never raw sensor data.
Session data and reporting stay scoped to your program — not pooled across unrelated organizations.
Reasoning is powered by Claude via the Anthropic API, under Anthropic's enterprise data-handling terms.
No. Face and voice processing happens locally in your browser. Only abstracted behavioral signals — not raw footage — are ever sent anywhere, and only for the purpose of generating real-time coaching.
A modern browser (Chrome recommended), a webcam, and a microphone. No install required — it runs entirely in the browser.
AdaptIQ started as a hackathon project (1st place, Claude NJIT Hackathon 2026) and individual practice sessions are free to use. Program-level plans below are for organizations that want reporting, admin tools, and support at scale.
Firebase handles authentication — we never see or store your Google password. Once signed in, you're taken straight into the interview practice product.
ADHD, Anxiety, and ASD-oriented profiles today, each with different intervention styles and sensitivity thresholds — you can also select specific weaknesses to focus on (eye contact, stuttering, distraction, pacing, anxiety).
Every plan includes profile-based coaching, session reporting, and privacy-first processing. Higher tiers add seats, admin controls, and support.
Sign in and you're straight into a mock interview session — or talk to us about rolling AdaptIQ out across your program.