A Senior Software Engineer role. 15 candidates. Our AI screening pipeline scored, ranked, and drafted personalised outreach emails for every shortlisted candidate — in under a minute. Here's exactly what it produced.
"This is a live demonstration using our production screening pipeline. These are simulated candidates but the AI processing is real."
15
CVs processed
<60s
Total processing time
4
Shortlisted (≥75)
~8 min
Recruiter time needed
The maths are brutal — and that's before accounting for screening fatigue.
| Task | Time per CV | 15 CVs | 50 CVs/day |
|---|---|---|---|
| Read and digest CV | 2–3 min | 30–45 min | 1.7–2.5 hrs |
| Compare against job spec | 1–2 min | 15–30 min | 50 min–1.7 hrs |
| Write notes / score | 1 min | 15 min | 50 min |
| Total | 4–6 min | 60–90 min | 3.3–5 hrs |
That's half a working day spent reading CVs
Before a recruiter has made a single call. And that's before accounting for the cognitive load — reading CV number 38 of 50 with the same attention as CV number 1 is genuinely hard. Screening fatigue leads to missed candidates and inconsistent assessments.
From job spec to ranked shortlist with outreach emails — fully automated.
Job spec ingested and parsed
Role requirements, stack, experience level, and culture signals extracted
Each CV submitted to scoring engine with job spec context
All 15 CVs processed in parallel against the same criteria
AI assesses each candidate across 5 weighted dimensions
Technical skills, experience level, industry relevance, cultural indicators, overall fit
Structured JSON output returned
Scores, flags, recommendation, and written summary for every candidate
Ranked shortlist generated automatically
Sorted by score, banded by recommendation tier
Personalised outreach emails drafted for all shortlisted candidates
Each email references specific CV details — not templates
| Metric | Result |
|---|---|
| CVs processed | 15 |
| Processing time | < 60 seconds |
| Shortlisted (≥75) | 4 candidates |
| Maybe (45–74) | 4 candidates |
| Rejected (<45) | 7 candidates |
| Outreach emails generated | 5 (top candidates + next best) |
| Recruiter time required | ~8 minutes (review + approve) |
The pipeline correctly identified the signal from the noise. 7 of the 15 candidates (47%) were unsuitable — including a marketing manager who had applied to the wrong role, a junior developer with 2 years' experience, and a finance professional exploring a career pivot. These were rejected in milliseconds, with a clear rationale.
| Step | Manual Process | ShortlistOps |
|---|---|---|
| Read all 15 CVs | 30–45 minutes | Instant |
| Score each candidate | 15–30 minutes | Instant |
| Create ranked shortlist | 15 minutes | Instant |
| Write notes on each | 15 minutes | Automatic |
| Draft outreach emails | 20–30 minutes | Automatic |
| Consistency | Variable (fatigue, bias) | Consistent, auditable |
| Output quality | Notes in a spreadsheet | Structured data, JSON, ranked table |
| Outreach personalisation | Often templated | Specific to each candidate's CV |
| Total recruiter time | 90–135 minutes | 8–10 minutes |
Time saved per 15-CV batch
80–125 minutes
Quality improvement
Every candidate assessed on the same criteria, with documented rationale
These are exactly what our pipeline produced — scores, summaries, flags, and recommendations.
"James is an exceptional match — a 7-year backend engineer with direct, hands-on experience in fintech payments infrastructure at Monzo and TrueLayer, using exactly the stack FinVault runs. The only minor consideration is salary expectations may be above the advertised range."
| Technical Skills Match | 97 |
| Experience Level | 95 |
| Industry Relevance | 98 |
| Cultural Indicators | 88 |
| Overall Fit | 93 |
Green flags: FPS integration at Monzo, idempotency layer design, open-source maintainer, PyCon speaker, exact stack match
Red flags: May exceed salary range; motivation for leaving should be explored
"Sarah is a capable senior Python engineer with a solid technical foundation — her stack maps well to requirements, and she has built event-driven reconciliation systems which is directly relevant. However, she has no fintech or financial services background, which is a meaningful gap for a Payments Core role."
| Technical Skills Match | 78 |
| Experience Level | 72 |
| Industry Relevance | 42 |
| Cultural Indicators | 80 |
| Overall Fit | 68 |
Recommendation: Maybe — consider if shortlist pool is thin
"Claire is a Senior Marketing Manager — an experienced, high-calibre professional in her field — but she has applied to a Senior Software Engineering role and has no software engineering background whatsoever. This application cannot be considered for this position."
Recommendation: Reject — redirect to any marketing roles if available
Recruiter time saved:
4–6 minutes of reading a CV that should never have reached the desk
| Rank | Candidate | Score | Recommendation |
|---|---|---|---|
| 1 | James Hartley | 93 | ✅ Shortlist |
| 2 | Alex Chen | 91 | ✅ Shortlist |
| 3 | Priya Nair | 88 | ✅ Shortlist |
| 4 | Sofia Mendes | 76 | ✅ Shortlist |
| 5 | Nina Kowalski | 71 | 🟡 Maybe |
| 6 | Sarah Okonkwo | 68 | 🟡 Maybe |
| 7 | Rachel Stone | 58 | 🟡 Maybe |
| 8 | Tom Bradley | 52 | 🟡 Maybe |
| 9 | Daniel Foster | 48 | ❌ Reject |
| 10 | Oliver Singh | 44 | ❌ Reject |
| 11 | Marcus Williams | 38 | ❌ Reject |
| 12 | Anita Patel | 22 | ❌ Reject |
| 13 | Ben Taylor | 11 | ❌ Reject |
| 14 | Jamie Cross | 9 | ❌ Reject |
| 15 | Claire Dobson | 3 | ❌ Reject |
Every shortlisted candidate gets a personalised message referencing their specific CV — written by AI, reviewed by a recruiter, sent in seconds.
Hi James,
I'm reaching out about a Senior Software Engineer opportunity that I think is a strong match for your background — specifically the payments infrastructure work you've been doing at Monzo.
The role is at FinVault Technologies, a Series B fintech processing £2bn/month in treasury, payments, and FX for mid-market businesses. They're building out their Payments Core squad, which owns the real-time payment processing engine, bank integrations, and reconciliation systems. Given your work on Monzo's FPS integration and reconciliation service, this is essentially the same problem domain.
A few things that stood out from your profile:
Stack-wise it's a very clean match: Python (FastAPI), Go, Kafka, PostgreSQL, AWS, Terraform, Kubernetes, Datadog.
Would you be open to a 20-minute conversation this week?
AI-generated · Ready for recruiter review · Sent with one click
Notice what this email does: it references specific details from James's CV — the idempotency layer, the 94% stat, fps-client, the PyCon talk. This isn't a template — it's a personalised message that will get a response. Written by AI, reviewed by a recruiter, sent in seconds.
For a recruitment agency screening 50 CVs per day.
3.7h
saved every day
vs. manual screening
75h
saved every month
per agency
£31,500
saved annually
at £35/hr consultant cost
| Metric | Manual | ShortlistOps | Saving |
|---|---|---|---|
| Time per CV (screen + notes) | 5 min | 0.5 min | 4.5 min |
| CVs per day | 50 | 50 | — |
| Time per day | 4.2 hrs | 25 min | 3.7 hrs/day |
| Time per week (5 days) | 20.8 hrs | 2 hrs | 18.8 hrs/week |
| Time per month | 83 hrs | 8 hrs | 75 hrs/month |
| Cost (at £35/hr) | Manual | ShortlistOps |
|---|---|---|
| Monthly recruiter hours on screening | 83 hrs | 8 hrs |
| Monthly recruiter cost | £2,905 | £280 |
| Monthly saving | £2,625 / month | |
| Annual saving | £31,500 / year | |
That's the equivalent of half a recruiter's salary, freed up for revenue-generating work.
And that's before accounting for:
When screening is automated, your team can focus on what only humans can do.
With candidates who actually want to hear from you — because your outreach is personalised.
Go deeper with clients — understand what really makes a great hire for them, not just the job spec.
Through the interview process — prep calls, feedback, managing expectations. The stuff that wins placements.
New client relationships, account growth, market mapping. Revenue-generating activity.
The edge cases the AI flags for human review — where your experience and instincts actually matter.
Recruitment is a relationship business.
ShortlistOps handles the processing. You focus on the relationships.
Job spec: Realistic specification for a Senior Software Engineer role at a fictional fintech company (FinVault Technologies)
Candidates: 15 fictional candidates designed to represent a realistic spread of applicant quality
Scoring: Performed by Claude (Anthropic's AI) using ShortlistOps's production screening prompt
Results: Real AI outputs — the scores, summaries, flags, and recommendations shown above are exactly what our pipeline produced
Emails: Generated by AI, ready for recruiter review and personalisation before sending
The candidates are simulated. The AI processing is real. The time saving is real.
Book a demo and we'll run your next real batch of CVs through the pipeline live on the call — your actual role, your actual candidates.
Or email us at anas@shortlistops.co.uk — we'll get back to you within 24 hours.
Call or WhatsApp: +44 7388 281312