Voice Features
Stori uses voice in three places — editing your profile, searching for hiring managers, and completing the AI interview itself. Each one uses voice differently, so the experience matches what you're trying to do.
Where Voice Works
You'll find a microphone icon in three places across the candidate product:
- Your profile page — voice-to-edit. Speak commands like "Add React to my skills" or "Update my summary" and Stori applies the change.
- Recruiter Source page — voice search. Describe who you're looking for and Stori parses it into filters automatically.
- Stori Interview — full voice agent. The AI asks a question, you answer, it asks the next one. No typing.
Voice-to-Edit Profile
Open your profile page and click the mic icon in the top right. Speak any of these commands:
- "Add React, TypeScript, and Postgres to my skills."
- "Update my summary to say I'm a Senior Engineer with eight years of experience in fintech."
- "Add my time at Stripe — Senior Engineer, 2021 to 2023, San Francisco."
- "Change my city to Austin."
- "Add my Stanford computer science degree, 2015 to 2019."
Stori transcribes what you said, an AI planner figures out what fields to update, and the change is applied. A toast notification confirms what was changed so you can verify.
Voice-to-edit supports 25+ action types across name, contact info, work history, education, skills, projects, certifications, languages, links, awards, and your summary.
Voice-to-edit is great for the messy first pass of building your profile. It's much faster to talk through your work history than to type it field by field. Then go back and tighten the wording afterward.
Voice Search on the Source Page
On the Recruiter Source page, click the mic icon next to the search filters. Describe what you're looking for in your own words:
- "VPs of Engineering at fintech companies in San Francisco."
- "Heads of Product at companies between 50 and 200 people."
- "Director of Marketing in Los Angeles."
Stori transcribes your query, an AI parses the components (job title, location, company size, industry), and the search runs automatically with the right filters applied. Faster than clicking through each filter dropdown.
The Voice Interview
The Stori interview itself is fully voice-driven. Built on OpenAI's Realtime API over WebRTC, it's a conversational back-and-forth — the AI asks a question, you answer naturally, it acknowledges and moves to the next question.
Key things to know:
- Latency — sub-200ms response times, so it feels like a real conversation, not a transcription tool.
- Audio quality — echo cancellation and noise suppression are on by default. A built-in laptop mic in a quiet room works fine.
- Free on every plan — the interview is free whether a recruiter invites you or you start it on your own.
- Re-recording — on your own, you can re-record your interview every 90 days.
What's Under the Hood
Stori uses two different stacks for voice depending on the use case:
- Whisper transcription for voice-to-edit and voice search. Single-pass transcription, then an AI planner interprets the command. No ongoing conversation.
- OpenAI Realtime API for the interview. WebRTC connection, bidirectional audio, conversational turn-taking. The AI hears you in real time and responds in real time.
Both run on OpenAI infrastructure. Stori never stores raw audio longer than needed for transcription; what's persisted is the transcript text.