Candidates join a live video interview with Hira, Hiearlly's AI interviewer. Hira talks with them in real time, asks follow-up questions based on what they say, and ends every interview with a report that scores each skill and quotes the answers behind the scores.
HiraYou mentioned the API was timing out. What did you check first?
Rounds
L1 screening and L2 technical or role-specific rounds
Format
A live, two-way video conversation with Hira
When
About 30 minutes, at a time the candidate books, on phone or laptop
Interviewer
Hira, who listens, responds and asks follow-up questions
Output
A report with a score per skill and transcript quotes
Language
English today, with more languages in development
How it works
L2 technical roundLinuxSQLREST APIsEscalations
Choose the round and skills
Pick L1 or L2 and the skills to assess. Hira builds the interview from your job description.
Nice work on the test, Riya. Pick a time for your interview.Book a slot
Candidates book a time
Invites go out on WhatsApp and email, with automatic reminders until the candidate books.
HiraWalk me through how you would debug a failing API call.
Hira interviews them live
A real-time video conversation. Hira follows up on each answer, and proctoring runs throughout.
91
Example
Ranked #1 of 214
Your team reads the report
Each candidate is scored and re-ranked, so recruiters start with the strongest.
Every score shows the answer behind it
Recruiters don't have to trust a number. Each report shows how the candidate answered, what stood out and what to ask in the next round.
91
Riya Menon
Technical support engineer, L2 round
Example
Riya explains API debugging step by step and backs it with a real incident from her current role. She is strong on Linux and SQL, and less experienced with customer escalations.
Communication
93/100
Problem solving
88/100
Linux
90/100
Escalations
62/100
Strengths
Checks logs, then network, then the client before guessing.
Worth probing
Has not owned a customer escalation end to end. Ask for an example in the next round.
Turn 8
I checked the gateway logs first, and the timeouts lined up with one database query that had lost its index.
Shows a structured approach to finding the cause, relevant to Problem solving.
A score for every skill
Each skill you chose gets its own score against the same rubric, so candidates are compared fairly.
Strengths and what to probe next
The report tells the next interviewer where to dig, so L2 doesn't repeat L1.
Quotes from the transcript
Key answers are quoted with the reason they matter, so anyone can check the score.
Integrity signals
Proctoring and fraud flags appear beside the report as information for your team. They never reject anyone automatically.
Sentiment analysis (coming soon)
Sentiment analysis of interview answers is coming soon.
An AI interview is a structured video interview run by software instead of a recruiter. In Hiearlly, Hira holds a live conversation with the candidate, follows up on their answers and scores each skill against the same rubric for every candidate.
Is it a live interview or a one-way video?
It is live. Candidates join a real-time video call with Hira, who responds to what they say and asks follow-up questions, the way a human interviewer would.
What is the difference between L1 and L2 interviews?
L1 checks that the candidate has the basic qualifications for the role. L2 goes deeper to find out whether they know what they are talking about and can prove it.
Does the AI reject candidates?
No. Hira scores and ranks candidates, and your team makes every decision. Integrity flags are information for recruiters, never an automatic rejection.
How long does an AI interview take?
About 30 minutes. Candidates take it at a time they book themselves, on a phone or laptop.
Which languages can Hira interview in?
English today, with more languages in development.
See an AI interview on one of your roles
Book a 30-minute demo. Bring a job description and we'll run a sample L1 interview for it.
