Your in-house recruiting agency.
From a role’s requirements to a screened shortlist[1] the same day, so you can do the human part, not the legwork.
[1]“Built ADF pipelines ingesting market data”Priya_Raman_CV.pdf, p. 1found in the file
- Strong signalJonas Berg2ndSenior Data Engineer · Nordlys Telecom
- Thin profileSofia Rossi3rdData Engineer II · Lumen Health
- Some signalRadu Ionescu2ndCloud Data Engineer · Orbis Retail
- Strong signalAisha Khan1stSenior Data Engineer · Fenwick Markets
- Azure Data Factory“Built ADF pipelines ingesting market data” · verified
- SQL Server“Owned the SQL Server warehouse end to end” · verified
- Databricks“Moved two batch jobs onto Databricks” · verified
- Payments or capital markets“Led payments integration for a card scheme” · not in the CV
Where a recruiter’s hours go.
The legwork Scout takes on is where recruiting time goes, by the industry’s own numbers.
- 17.7h
hours of admin per vacancy, 3.6 of them spent reviewing applications.
Totaljobs, survey of 748 UK HR leaders, 2025 - 20%
of the working week saved by recruiting teams using generative AI. About a day a week.
LinkedIn, Future of Recruiting 2025, 1,271 recruiting professionals - 5×
more likely to be hired when a candidate was sourced than when they applied.
Gem, 2025 Recruiting Benchmarks, 140M+ applications
Every score shows why.
Scout reads each CV against the role’s rubric instead of matching keywords. It quotes the CV for every criterion it marks as met and drops any quote it can’t find in the file. Open a score to see the quotes behind it.
- Built ADF pipelines ingesting market data from fourteen vendors into one warehouse.
- Owned the SQL Server warehouse end to end, from modelling to query tuning.
- Moved two batch jobs onto Databricks as a pilot for the risk team.
- Wrote Python services for trade reconciliation and end-of-day reporting.
- Ran the team’s on-call rota for the reporting stack.
- 1CheckingAzure Data Factory20must“Built ADF pipelines ingesting market data”
- 2CheckingSQL Server20must“Owned the SQL Server warehouse end to end”
- 3Checking5+ years in data engineering15must“Data Engineer, Halden Labs, 2018 – 2021”
- 4CheckingPython10must“Wrote Python services for trade reconciliation”
- 5CheckingDatabricks10“Moved two batch jobs onto Databricks”
- –CheckingPayments or capital markets10“Led payments integration for a European card scheme”
- –CheckingTerraform5No quote offered
Quotes are checked
Met or partly met needs a quote from the CV. If Scout can’t find it there, the claim is dropped.
A 75 is always a 75
The AI checks each criterion against the CV. A fixed formula then turns those checks into the score, the same way on every role.
No verdicts from profiles
LinkedIn profiles are sorted into strong, some or thin. Only a scored CV can be marked red, and a dealbreaker needs a quote of at least six words.
How a role moves through Scout.
Most roles go in this order, and finishing one step starts the next in the background.
- JD1/6
Paste what the hiring manager sent.
Text, a PDF or an email, in any language. Scout writes the JD in English and marks any gap the brief leaves for you to fill.
- Notes from the intake call override the written brief
- A finished JD is used unchanged
Senior Data Engineer · job descriptionTranslated from RomanianWay of workingHybrid · 3 daysLondon, UKThe roleMeridian Financial is building a new market-data platform on Azure. You will own the pipelines that feed it, from ingestion to the warehouse.
What you will do- Design and run Azure Data Factory pipelines for market and reference data
- Own the SQL Server warehouse, its models and its performance
- Work in a team of team size? engineers and analysts
Contractcontract type?Permanent1 gap to fill · 2 facts inferredAnswer gapsWhat the hiring manager said“I won't look at anyone without payments experience.”
Overrides the written brief in the rubric - Brand2/6
JDs on your letterhead.
Upload a JD you’ve already published. Scout copies its colours, footer and headings for every new one.
- The preview on screen is the exported PDF
- Diacritics print correctly, so București stays București
Meridian FinancialSenior Data EngineerHybrid · London · PermanentThe roleWhat we are looking forWhy MeridianMeridian Financial SRL · Strada Lipscani 12, BucureștiRead from your last JDBanner#123C3AAccent#E3A64BPaper#F4F1EAFooterHouse styleSection labelsLogo and banner, attached by you - Search3/6
A rubric and searches for each role.
Scout writes the screening rubric and the LinkedIn boolean searches in the background. You can edit both before anything is scored.
- Dealbreakers are worded as the reason someone is ruled out
- Each edit saves a new version of the rubric
Screening rubricv3 · activeMust haveAzure Data Factory20SQL Server205+ years in data engineering15Python10Nice to haveDatabricks10Payments or capital markets10Terraform5DealbreakerCannot attend the London office three days a week and is unwilling to relocateLinkedIn searches4 written("data engineer" OR "data platform engineer") AND ("data factory" OR adf) AND ("sql server" OR t-sql) NOT (junior OR intern)
2 · ("azure data engineer" OR "cloud data engineer") AND …
- Sourcing4/6
Sourcing in your own browser.
The Scout collector, a Chrome extension, works through the searches in your LinkedIn session at about the pace a person would.
- Pause or stop it at any point
- It stops at a captcha, a sign-in page or a warning
linkedin.com/search/results/people/?keywords=data%20engineerScout collector18of 40 collectedRunning · your account- Typed search 2 of 4
- Set filter · Greater London
- Read results page 1 · 25 people, 3 already on the role
- Left out Daniel Okafor · works at Meridian Financial
- Opened Priya Raman · next open in 90 s
PauseStop - Candidates5/6
Candidates screened as they arrive.
Each profile is labelled strong, some or thin against the rubric, with a message drafted from your templates.
- You send each message yourself from LinkedIn
- Export the list to Excel or CSV
Candidates · 38Export- Priya RamanStrong signalDraft ready
- Aisha KhanStrong signalSent · 2 days ago
- Tomáš NovákSome signalDraft ready
- Lena HoffmannThin profileNot drafted
Draft for Priya Raman · template “First message”Hi Priya, I'm hiring a Senior Data Engineer for a fintech building a new market-data platform in London. Your pipeline work at Northwind Capital stood out. Open to a short call this week? Best, Ioana
CopyMark sentFilled from your template - CV scoring6/6
CV scores with the evidence shown.
Drag CVs from email, WhatsApp or LinkedIn onto the role. Each one is scored as above, quotes and all.
- Green, amber or red against the rubric’s thresholds
- Drop a CV straight onto the candidate it belongs to
Drop CVs from email, WhatsApp or LinkedIn on the roleShortlist · ranked by score9 CVs- 1LinkedIn84greenAisha Khan6 of 7 criteria evidenced
- 2Email78greenPriya Raman5 of 7 criteria evidenced
- 3WhatsApp67amberIoana Marin5 of 7 criteria evidenced
- 4Email61amberJonas Berg4 of 7 criteria evidenced
- 5LinkedIn34redMateo GarcíaDealbreaker: cannot attend the office
Quotes checked against each CVOpen evidence
Meridian Financial is building a new market-data platform on Azure. You will own the pipelines that feed it, from ingestion to the warehouse.
- Design and run Azure Data Factory pipelines for market and reference data
- Own the SQL Server warehouse, its models and its performance
- Work in a team of team size? engineers and analysts
“I won't look at anyone without payments experience.”
Use it alone or with your team.
Each account starts as a private workspace. If you start a team, roles created in it are shared with your colleagues, along with their sourcing sessions, candidates, CVs and scores.
See who is on which step
The role header shows which colleagues have it open and which step they’re on, so two people don’t work the same candidates.
An activity feed
See who imported candidates, who marked messages as sent and whose CVs arrived. The feed is built from the role’s own records, so it always matches them.
Shared and private templates
A template can belong to the team or to one person. The {{recruiter_name}} field fills in with the name of whoever creates the draft.
Pay for the work Scout does.
Scout is priced on usage: the roles it sets up, the profiles it screens and the CVs it scores.
- Role set up
The JD, the rubric and the LinkedIn searches, written from one brief.
Price not set yetper role - Profile screened
A LinkedIn profile read against the rubric as it arrives.
Price not set yetper profile - CV scored
A CV scored against the rubric, each quote checked against the file.
Price not set yetper CV - Match
A CV matched against your open roles, or a JD against the people you already have.
Price not set yetper match
- Role set up1 ×Price not set yet
- Profile screened240 ×Price not set yet
- CV scored18 ×Price not set yet
- Match6 ×Price not set yet
Each line is a unit from the price list. The counts are an example, not an average.
Common questions.
Most are about your LinkedIn account and what happens to candidates’ data.
Does Scout scrape LinkedIn?
No. Scout’s servers don’t connect to LinkedIn at all. Sourcing runs in your own browser through the Scout collector extension, using your LinkedIn session, so the requests come from your computer.
Is my LinkedIn account at risk?
Yes, there is some risk. LinkedIn’s User Agreement prohibits software that collects profile data, and the collector runs on your account. Scout’s terms say so, and you accept them before your first run. To keep the risk down, runs open about 40 profiles an hour by default and stop at the first captcha, sign-in page or warning.
Will Scout message candidates for me?
No. Scout drafts a message for each candidate from your templates. You copy it, send it from LinkedIn and mark it as sent. The only thing a run can send is a connection request without a note, and only in sessions where you’ve allowed it.
What can the hiring manager’s brief look like?
Pasted text, a PDF or a Word file, in any language. Scout writes the JD in English and tells you if it translated anything. If the hiring manager has already signed off a JD, you can use it as it is.
Do I need the Chrome extension?
It’s the quickest way to source, but you can work without it. You can run the searches by hand and paste the results back, import a CSV, or drop in CVs.
Where are CVs stored?
In the EU. CV files, the text read from them and everything else on a role are stored in Frankfurt (Supabase, on AWS eu-central-1), and a file opens only for people who can see the role or talent pool it belongs to. Some of the work happens in the US: the app runs on Vercel, and to read and score a CV its text is sent to OpenAI’s API. OpenAI doesn’t train its models on API data and keeps requests for at most 30 days.
What about GDPR?
Your company is the controller of candidates’ data, and Scout processes it on your behalf. It only holds the people and CVs you add, and only you, or your team for team roles, can see them. Scout doesn’t decide anything about a candidate on its own. A person on your team makes every call, and each score shows the CV quotes it rests on, so you can explain how someone was assessed. You can delete a CV from the talent pool, a sourcing session with its candidates, or a closed role with everything on it, files included.
Does the EU AI Act apply to Scout?
Yes. The AI Act lists AI that filters applications or evaluates candidates as high-risk (Annex III), and the Digital Omnibus moved those obligations from August 2026 to 2 December 2027. Among other things, those rules ask for human oversight, transparency and record-keeping. In Scout, a person reviews every result and makes the decision, each score shows the evidence behind it and is worked out the same way on every role, and each score records the model and prompt version used. Companies using Scout have duties of their own as deployers, such as telling candidates that AI helps screen them.
Try it on a live role.
Paste in a brief you’re working on today. Scout writes the JD, the rubric and the searches from it.