EchoHire
AI voice interviews · LLM analysis · ICP matching

Screen every candidate.
Hear the ones that matter.

EchoHire sends candidates a link, conducts a real voice interview with AI, then scores, summarizes, and ranks them against your ideal candidate profile — automatically.

10–15 min
per AI interview
6 scores
every candidate
100%
transcribed & recorded
Pass / Fail
with reasoning

From invite to verdict in three steps

01

Send a link

Create a candidate and EchoHire emails them a unique interview link. They open it anywhere, any time — no scheduling, no apps.

02

AI conducts the interview

A natural ElevenLabs voice agent runs the conversation — background, skills, visa status, salary expectations, and your custom questions.

03

Get the full picture

Transcript, recording, structured profile, six interview scores, strengths and risks, ICP match percentage, and a pass/fail recommendation.

Built like the screening system you wish you had

Enterprise architecture under a recruiter-friendly surface — role-based access, monitoring, and your choice of analysis model.

Real voice conversations

ElevenLabs Conversational AI interviews candidates with natural turn-taking, recorded end-to-end.

Structured extraction

Location, visa status, education, experience, skills, and work preferences pulled into clean database fields.

Six-dimension scoring

Communication, confidence, clarity, professionalism, technical knowledge, and overall — each with written reasoning.

ICP matching engine

Define requirements and weights once. Every candidate gets a match percentage, gaps, risks, and a recommendation.

Ranking & analytics

Funnel, pass rates, score distributions, and a ranked candidate list your whole team can act on.

Enterprise-grade

Role-based access, rate limiting, webhook signature verification, Sentry and PostHog wired in.

One rubric. Every candidate.

Define your ideal candidate once — communication threshold, work authorization, LinkedIn, IT experience, relocation — and every interview is judged against the same bar, with hard-requirement failures and risk factors called out explicitly.

Your analysis, your model

Powered by OpenRouter: switch between Claude, GPT, Gemini, or DeepSeek models from the admin dashboard — no redeploys. Test connectivity with one click.

Claude SonnetClaude OpusGPT-4oGemini 2.5DeepSeek V3

Stop scheduling screens. Start hearing signal.

Sign in to create your first candidate — the interview link does the rest.

Open the dashboard