Superficial Screening
The 6-second average résumé scan leads to talented candidates being filtered out before anyone ever talks to them.
AIVIEW turns a job description into a tailored, scenario-based interview that adapts to every candidate in real time — replacing static multiple-choice tests and one-way video screens with immersive simulations and defensible scoring.
Tell me about a time you had to defuse a tense customer situation. What did you do?
I lowered my voice, asked them to share their goal, and proposed a refund plus a follow-up call…
Recruiters lose weeks to manual screening while top candidates walk away from impersonal, biased, surface-level assessments.
The 6-second average résumé scan leads to talented candidates being filtered out before anyone ever talks to them.
Implicit biases routinely compromise the fairness of first-round hiring decisions, even when interviewers mean well.
Static MCQ tests and coding puzzles miss real hands-on capability — communication, judgment, prioritisation and conflict resolution stay invisible.
Lengthy, unengaging multi-stage processes drive away exactly the high-performing candidates the company most wants to hire.
A dual-portal, AI-powered interview platform that lets HR generate tailored simulations from a single job description and gives candidates an immersive, real-world assessment experience.
Generates tailored, scenario-based interview flows simply from a job title and description — review, edit, and approve before sending invitations.
Delivers immersive role-play simulations — replying to client emails, prioritising tasks, working through real workplace scenarios — instead of generic, static questions.
An AI-powered platform that adjusts scenarios in real time based on candidate performance — branching, escalating, or correcting on the fly.
Evaluates responses in real time to escalate challenges for high performers or introduce corrective stages for those who need them — every candidate gets the interview that suits them.
Uses modular components — multiple choice, email, task prioritisation, role-play — grounded in the realistic workplace challenges of the specific role being hired for.
Acts as an AI-assisted tool where HR retains final authority — flows are reviewable and editable, scoring is transparent and defensible, ensuring human oversight and fairness.
Two of the screens candidates and HR actually see — built straight from the React frontend.
Candidates respond to a live customer email, draft a reply, and the platform evaluates tone, clarity, problem-solving and ownership in real time — feeding the adaptive engine.
Hi,
I placed this order nine days ago, paid extra for express shipping, and nobody has bothered to email me back. This is unacceptable. I want a real answer or a refund, today.
— Marcus
Per-candidate dimension breakdown, the full conversation timeline, and a visualisation of how the interview actually adapted — escalations, remedial branches and final outcome.
From AI-driven interview generation and adaptive branching, to modular workplace simulations, voice role-play, and a defensible rubric — every capability is shipped.
The brain of the platform — evaluates every answer in real time, escalates difficulty for high performers, and inserts up to three corrective remedial stages for those who are struggling.
HR pastes a job title and description; the LLM produces a complete, tailored interview flow with the right mix of components for the role — ready to review, edit and send.
Goes far beyond multiple choice — email simulations, task prioritisation boards, open-ended scenarios, and persona-driven role-play, all in one consistent interface.
The LLM plays realistic characters — an angry customer, a curious manager, a technical teammate — to test how the candidate responds to actual workplace pressure.
Recruiters review every AI-generated stage, edit individual questions, swap components, then approve the flow before candidate-specific secure invitations are sent.
HR sees per-candidate dimension breakdowns, the full conversation timeline, the actual adaptation path the interview took, and a downloadable evaluation report.
The right mix is auto-selected for each role — a developer gets coding-context tasks, a support agent gets email scenarios, an ops candidate gets prioritisation boards.
A defence-in-depth pipeline — deterministic no-answer floor → strict score-band rubric → weighted-average sanity check — produces evidence-based scores instead of inflated guesses.
User authentication, candidate-specific secure invite links, privacy-preserving session storage, and a multi-layer defence against prompt injection.
STT/TTS-powered live voice role-play with a chosen interviewer persona, for roles where verbal communication is the most important signal — same strict rubric as text mode.
Five tightly-coupled components work together to turn a candidate's conversation into structured, defensible scores.
Candidate logs in, browses positions, uploads a CV and starts the assigned interview in voice or text mode.
React · MUI · STT/TTS
Each answer streams to the AI Engine over WebSocket / REST — voice replies are transcribed live with Whisper.
WebSocket · Whisper · TTS
The LLM role-plays a persona, picks the next question's type and difficulty, and grades the previous answer against HR's rubric.
FastAPI · OpenAI · Pydantic
A defence-in-depth pipeline (deterministic floor → strict rubric → weighted-average sanity check) produces dimension scores and a recommendation.
Python · Strict rubric
HR sees the full conversation transcript, per-dimension chart, recommendation, and downloadable evaluation report.
Analytics · Reports
A modern, production-grade stack powering every layer of the platform.
A two-minute walkthrough covering candidate flow, adaptive questioning, voice role-play, and HR analytics.
Recruspace
