Content pipeline
Content pipeline running on autopilot
Keyword research and publishing pipeline for an in-house SEO stack. Indexed pages generated, linked, and maintained without a writer on payroll.
Publishing without an in-house writer
Recruitment
Consultants stop sourcing at 11pm, the pipeline does it by 7am
Every open role gets a nightly sweep across LinkedIn, GitHub, and niche directories. Candidates are enriched, scored for fit, and a shortlist of 20 lands on the consultant's desk every morning.
Client
UK tech recruitment agency
Build time
6 weeks to production
Headline outcome
Sourcing without a research team
The problem
Recruitment agencies are limited by consultant hours. Finding 10 viable candidates for a niche role takes 3-5 hours across LinkedIn Recruiter, Boolean searches, and portfolio sites. Consultants either skip sourcing for easier roles or eat into their evening. LinkedIn Recruiter licences cost £8-12k each and still require manual work.
The system
A per-role sourcing engine. Each role has a structured profile (skills, seniority, stack, company-pattern, location). The engine scans LinkedIn, GitHub, Stack Overflow, niche dev directories nightly, enriches each match (current company, recent activity, open-to-work signals), and scores by fit. Shortlist of 20 lands in Bullhorn or JobAdder every morning.
How it runs
The workflow, end to end.
01
Role profile extraction
Each open role has a structured profile pulled from the JD: must-have skills, seniority, stack, target companies, location, comp range.
02
Multi-source candidate discovery
LinkedIn (authenticated), GitHub, Stack Overflow, niche directories are scanned nightly. Candidates are de-duplicated across sources.
03
Enrichment + signal detection
Current role, tenure, recent activity, open-to-work signals, portfolio links are pulled per candidate.
04
Fit scoring
Candidates scored against the role profile on must-haves, nice-to-haves, and signal strength. Score + reasoning attached.
05
ATS handoff
Top 20 per role land in Bullhorn / JobAdder tagged to the job with a one-line pitch and the full profile attached.
Outcomes
What it moves.
500+
Candidates scored / night
-18
Consultant sourcing hours / week
+55%
Shortlist-to-submit rate
3
LinkedIn Recruiter seats replaced
Stack
What it's built on.
- Apify (LinkedIn + directories)
- GitHub API
- Bullhorn / JobAdder API
- Claude (scoring + pitch)
- Postgres + Trigger.dev
RecruitmentSourcingScreening
Running 20+ live roles? Let's scope the sourcing engine.
20 minutes. No slides. One process, one prototype.
Book a 20-min introMore builds
View all