Screen these 3 candidates before I interview them
Your AI CEO walks into every interview already knowing the candidate.
You'd think this needs 90 min per candidate of resume reading — Astra hands you one-page briefs while you finish coffee.
The short answer
Astra screens candidates by reading everything publicly available about them and synthesizing a one-page brief tailored to your role. For each candidate she pulls the resume, LinkedIn profile, GitHub (for engineers), public writing or talks, and any past company context, then outputs: top 3 reasons to hire, top 2 risks or red flags, 5 tailored interview questions targeting the candidate's specific gaps and strengths, and one "deep cut" question grounded in something they shipped or wrote. She cross-references their background against your role spec (pulled from your job post or recent chat) and flags any mismatch (e.g., "role says principal-level architecture, last 3 jobs were IC implementation"). Output is one Lark message per candidate, ready to read on your phone before the call. 5 minutes of prep replaces 90.
How Astra actually does it
- 1Gather sources per candidate
Resume PDF, LinkedIn profile, GitHub (repos + contributions), Twitter/X if relevant, blog posts, conference talks, prior companies' Crunchbase entries.
LinkedIn Sales NavigatorWebSearchGitHub - 2Pull the role spec
Read the job post, prior chat about what you're hiring for, and your hiring rubric if you have one in Notion.
- 3Synthesize the brief
One page per candidate: 3 hire signals, 2 risks, role-fit score 1-10, 5 tailored questions, 1 deep-cut question grounded in something they shipped.
- 4Flag mismatches
Title inflation, gaps in employment, repeated short tenures, stale tech stack vs your needs, or anything that warrants follow-up in interview.
- 5Deliver in Lark
One message per candidate, formatted for phone reading. Linked sources at the bottom. Sent at least 30 min before each interview.
Lark
What it looks like in chat
One Lark message per candidate (5-7 bullets, phone-readable) with hire signals, risks, role-fit score, 5 tailored questions, and a deep-cut question. Source links at the bottom.
On-demand per interview batch.
Ask Astra this right now
We'll spin up your workspace, hand the prompt to Astra, and you see the answer in 60 seconds. Free.
Try this with AstraFrequently asked questions
What if the candidate has a sparse online presence?
Astra reports what she found honestly — "low public footprint, brief based on resume only" — and tilts her interview questions toward open-ended exploration rather than pre-loaded specifics. She never invents details to fill gaps.
Can Astra also rate candidates or pick a winner?
She gives a role-fit score 1-10 per candidate but explicitly avoids ranking — hire decisions involve fit, team, and gut signals she can't see. Her job is to make sure you walk in informed; the call stays yours.
What about confidentiality — does she store the resumes?
Resumes are processed in-session and discarded. They're not added to any training data or persistent memory. The brief itself stays in your Lark thread, scoped to your workspace, and you can delete the thread anytime.
Can she do this for 10+ candidates?
Yes — she scales linearly, ~8-12 minutes per candidate to gather sources and write the brief. For a batch of 10 you'd have all briefs in ~2 hours. She'll prioritize by your interview schedule so the next-up candidates land first.
Run your one-person company.
Hire your AI team in 30 seconds. Start for free.
Free to start · No credit card required · Set up in 30 seconds