Compass is AI-powered search across the candidates you already have. Point a live job spec at your CV database and instantly surface qualified people from past applications, unlocking value you've already paid for, before spending on new adverts.
Every application you have ever paid to attract is sitting in your database, and most of it will never be looked at again. When a new role opens, the reflex is to advertise, because searching thousands of old CVs with keyword tools is slow and unreliable: the right person is in there, filed under a job title that doesn't match the search box. So agencies and TA teams pay, again, for candidates they already have. Compass exists to end that reflex.
Search when a role opens, match when a candidate arrives, and keep the results working for you
Point a live job spec at your database. Compass reads the requirements, casts a wide net across past applications, and returns a ranked shortlist where every candidate carries per-requirement evidence for why they fit. Seconds, not an afternoon of keyword guessing.
A strong candidate applies for a role that isn't quite right. Instead of a rejection and a lost relationship, Compass shows your recruiters every open role that candidate does fit, ranked the same evidenced way. Nobody good leaves the funnel just because they arrived through the wrong door.
Searches become living shortlists saved against each role, so sourcing work is never thrown away. The pipeline you build for one vacancy is still there, evidenced and ranked, when the client calls about the next one.
Most AI search tools rank candidates by a similarity score nobody can explain. Compass doesn't. Retrieval only casts the net; the ranking is done by scoring each candidate against each requirement individually, with the supporting evidence recorded, exactly the scorecard approach behind CloudFlow's screening engine. A requirement can be met, partial or missing, weights are applied in the open, and when a client asks why a candidate was suggested, your answer is a line-by-line breakdown rather than a mystery percentage. The full methodology is written up in our engineering article on candidate scoring.
The same rules that govern CloudFlow screening govern Compass
Name, age, gender, nationality, ethnicity and photos are never used in any scoring path. Candidates are ranked on evidence of skills and experience, nothing else.
Requirements filter on meets-or-exceeds: a candidate with more experience than the role asks for is flagged, never down-ranked. Exceeding the bar is treated as clearing it.
Where a role states eligibility requirements, Compass filters on candidates' right to work in that jurisdiction, so the shortlist is one your team can actually act on.
Every client's candidate data is searchable by that client alone, enforced at the infrastructure level, and hosted in the EU. Your database is never part of anyone else's search.
Each search records what was asked, which filters applied and how every candidate scored, giving you a full decision trail for clients, auditors and GDPR accountability.
When a candidate asks to be erased, everything goes: the CV, the derived search data, all of it. Erasure means erasure, in line with your GDPR obligations and ours.
Compass is an optional add-on to the CloudFlow platform. It works with the ATS and HR systems you already run, alongside the screening engine that reads every new application as it arrives. New applications get screened; the database they accumulate into gets searchable. Together, no application you have ever received is wasted.
Ask us for a Compass demo on your own roles. Bring a live job spec; we'll show you what your database has been hiding.