From a flood of CVs to a structured, trackable pipeline.
ClientA leading overseas recruitment & manpower agencySectorOverseas employment & manpower
High-volume recruitment meant thousands of CVs in inconsistent formats and a candidate pipeline tracked across spreadsheets and inboxes. We built a system that parses documents into structured records and moves candidates through a clean, automated pipeline.
recruit.internal/pipeline
Applied128
CAND-4821DriverCV parsed
CAND-4822WelderCV parsed
Screening46
CAND-4790ElectricianDocs ✓
CAND-4788MasonInterview set
Medical23
CAND-4771PlumberGAMCA fit
Deployment12
CAND-4755TechnicianVisa ✓
Recruitment pipeline — synthetic data
0s
CVs parsed into structured records
0
source of truth for candidate status
0
systems unified (ERP, HRIS, portal, messaging)
01The challenge
Applications arrived as PDFs, images and forwarded messages — every one different. Recruiters re-keyed the same details into multiple systems, candidates fell through the cracks between stages, and nobody had a reliable, real-time view of where anyone was in the process.
02What we built
01Built a document-parsing pipeline that extracts structured fields from CVs and application documents automatically.
02Standardised candidates into one record shape feeding the ERP and HRIS — no more re-keying across tools.
03Modelled the recruitment pipeline as clear stages with automated movement, reminders and hand-offs.
04Connected messaging channels so candidates get timely, automated updates without staff chasing each one.
05Surfaced the whole pipeline on a live board so managers see bottlenecks the moment they form.
03Outcome
Documents become structured records automatically instead of being re-typed.
Every candidate has one status, visible in real time across the team.
Routine follow-ups and stage moves happen on their own.
Redacted screenshot