How Scout works

From the JD to the brief.

Scout sits on one step — the screen — and does it completely: capture the conversation, separate the speakers, map the evidence to your requirements, and hand back a brief. Here is the whole pipeline, end to end.

The workflow

Six steps, mostly automatic.

You scope the role and run the call. Scout handles everything between the conversation and the brief.

1

Scope the role

Paste the JD and any notes from the hiring manager. Scout drafts the topics to cover — yours to edit before the call.

JD · hiring-manager notes → topics to cover
2

Run the call

Take the intake call exactly as you do today. Scout listens silently and never prompts you mid-conversation.

Speakerphone · Zoom · Teams
3

Capture the audio

Your browser mic records the room, or a meeting bot captures each speaker on Zoom or Teams. You pick per screening.

Local-mic (default) · meeting bot (opt-in)
4

Transcribe & separate speakers

The recording becomes a speaker-labelled transcript, normalized to one schema regardless of how it was captured.

Speaker-separated transcript
5

Extract the signal

Scout reads the transcript against your topics and pulls the candidate’s own words for each one — covered, partial, or not mentioned.

Evidence mapped to your requirements
6

Review & brief

A few minutes after you hang up, the brief is in your vault. Render your verdict in The Read and export or paste it anywhere.

Summary · profile · coverage · your verdict
Capture

Two ways to capture. One pipeline.

How Scout listens is the only thing that changes between a speakerphone call and a headphones call. You choose per screening — Scout never auto-detects or joins a meeting you didn’t send it to.

Default · local-mic

On speakerphone

Take the call on speakerphone with Scout open beside you. Your browser microphone records the room; nothing joins the meeting. Audio is written locally during the call and uploaded in the background.

Opt-in · meeting bot

On headphones

On Zoom or Teams, Scout sends a meeting bot that joins as a participant and captures each speaker on a separate track. Same vault, same brief — cleaner speaker separation.

On speakerphone
Speakerphone
The call you run
Browser mic
Records the room
On device
Buffered locally
Storage
Encrypted audio store
On headphones
Zoom / Teams
The call you run
Meeting bot
Joins as participant
Per-speaker
Separate tracks
Storage
Encrypted audio store

Both paths land in the same store and converge on one pipeline.

Transcribe & extract

Speaker-separated transcript, then signal.

The recording is transcribed into a speaker-labelled transcript, normalized to one schema, and read for signal — what the candidate said about each topic, in their own words. The brief reads the same whether the call came in on speakerphone or through the bot.

Audio
From either path
Transcribe
Speaker-separated text
Normalize
One common schema
Extract
Scout reads for signal
Brief
Summary · profile · coverage

No signal logic depends on a field only one capture mode produces.

The brief

What lands in your vault.

A screening summary, a structured candidate profile, and topic coverage — each topic marked covered, partially covered, or not mentioned, with the candidate’s own quotes attached.

scout.vantumworks.io / screenings / sample

Screening summary

Software engineer with 8 years of experience. Has led a 6-person payments team at a fintech for the past 4 years; the team owns the routing layer for the transaction-processing platform. Resume lists Java and Python as production languages. The JD references Go as the primary language; the candidate stated they have written Go for side projects but not in production.

Topic coverage

Covered
5+ years distributed systems experience
“We were processing about 40 million transactions a day across three regions, and I owned the routing layer end to end.”
Partially covered
Experience with the payments domain
“Mostly card-present and ACH. Never touched cross-border or FX — that was a different team.”
Not mentioned
Go in production
Not raised on the call. Resume lists Go under side projects only.
The Read — review loop

Scout lays out the facts. You make the call.

Each requirement, the candidate’s own words next to it, and a verdict that’s yours. Scout never decides — it makes deciding fast, with the evidence already beside every call.

scout.vantumworks.io / screenings / sample / review
Your criteria
Scout found
Your call
5+ yrs distributed systems
“40M transactions a day across three regions; owned the routing layer end to end.”
Payments domain
“Card-present and ACH. Never touched cross-border — that was a different team.”
Go in production
Not mentioned on the call. Resume lists Go under side projects only.
Campaigns — resumes at scale

Post the role. Read every resume the same way.

Spin up a public page for a role and share the link. Resumes land in a queue — flip through them, render each one for free, and analyze the ones worth a closer look. Analysis is something you trigger, never automatic, so you decide what gets read.

scout.vantumworks.io / campaigns / senior-backend
mentions Kubernetesstates 5+ yrs backendUS-remoteOrder — most recent
A. Riveraapplied 2h ago
Scout found — States 8 yrs backend · mentions Kubernetes, Go · led a 6-person team
M. Oseiapplied 5h ago
Scout found — States 5 yrs backend · mentions AWS, Python · Kubernetes not mentioned
J. Parkapplied yesterday
Not analyzed yetReview →
L. Nguyenapplied yesterday
Not analyzed yetReview →

Every resume you analyze is held to the requirements you set and described in the candidate’s own words — same rubric, no fatigue, no drift. Campaigns come with hiring Rosa and with your free trial.

Scope, on purpose

What Scout does — and deliberately doesn’t.

Scout owns the screen and hands off neatly to everything you already use. The focus is the point: it keeps you out of yet another platform to manage.

Scout does this
  • Captures the intake calls you run — speakerphone or meeting bot
  • Reads every resume on a public role page against your requirements
  • Runs full intake conversations on your behalf when you hire Rosa
  • Sets your criteria beside the evidence so your verdict takes seconds
  • Keeps a private, searchable vault of everything you’ve screened
  • Exports to PDF and copies cleanly into any ATS
Scout doesn’t do this
  • Score, rank, or recommend candidates
  • Manage pipeline stages or run a Kanban
  • Send outreach, schedule calls, or chase candidates
  • Integrate with your ATS — you export and paste
  • Host a candidate database or marketplace
  • Coach you in real time during the call
  • Surface anything about you to your firm’s leadership
Privacy & trust

Your data stays yours.

Recruiters own their vault. Candidate data has a narrow purpose: the originating recruiter’s screening, nothing else.

Vault

Your vault, your data

Everything you screen lives in a private vault you control. Export it or delete it whenever you want.

Retention

Audio expires

Recordings are retained for 90 days, then removed automatically by lifecycle policy.

Purpose

Narrow purpose

A candidate’s data is used only for the originating recruiter’s screening — never pooled or resold.

Access

Scoped to you

Every screening is tied to your account — other recruiters never see your candidates or your reads.

Scout describes; you decide. Scout produces no scores, no rankings, and no recommendations about candidates — by design. It surfaces what was said and maps it to your criteria; the judgment stays with the recruiter.

See the brief before you sign up.

Walk through a real, structured brief — then start your free trial when you’re ready.

How Scout works