Carerix has been building recruitment, uitzend and detachering software for the Benelux since 2004, and it covers ground Kepler does not go near: contracts, timesheets, invoicing, client and candidate portals, and AVG/GDPR lawful basis as real data. Kepler covers less of the desk and more of the database — one assistant across everything you own, matching with the evidence attached, and one published price.
If you run a Dutch or Belgian secondment or temp agency, buy Carerix. Their detachering edition generates contracts, handles timesheet registration and drives invoicing from the same system that does the recruiting, they integrate the Benelux back offices, they model AVG/GDPR lawful basis with expiry dates and anonymisation, and they hold ISO 27001 — and Kepler has none of that, so this is not a close call. Kepler is the better product for a permanent or contingency agency whose real problem is a database nobody can search: one assistant that reads every CV, note, email and call transcript you own and can act on what it finds after you approve, matching with a verdict and the supporting line for every requirement, and one published price of USD 119 per user per month with every feature switched on.
Last checked 6 Sep 2026 · Written from Carerix's own public documentation and pricing pages. Tell us if we have something wrong and we will correct it.
Carerix Private AI is deliberately fenced: their own page says the system is scoped to candidate, vacancy and matching support and that general questions return an error. Kepler’s assistant sits in a sidebar on every page, reads CVs, notes, emails and call transcripts across the whole workspace, and can create and update records once you approve.
Kepler ranks candidates against criteria you wrote for that specific job, and every requirement expands into a pass or gap verdict with the line from the candidate’s record that supports it — so a consultant can argue the order with a client instead of quoting a number.
Carerix sells four editions and does not publish a price for any of them. Kepler is USD 119 per user per month billed annually, one edition, every feature included, unlimited records, migration done for you, and additional AI capacity available on request.
Dimensions where Kepler is the stronger product. Carerix is ahead of us on a great deal else — back office, timesheets and invoicing, portals, AVG/GDPR lawful basis, SMS, VoIP, multiposting, a mobile app and ISO 27001 — and the blog post covers all of it.
| Kepler | Carerix | |
|---|---|---|
| Asking your database a question | ||
| Scope of the AI assistantWhether it can answer a question that spans the database. | App-wide sidebar with about fifteen tools, reading the whole workspace | Assistant fenced to candidate, vacancy and matching support; general questions return an error |
| What it does with the answerReading is half the job. | Returns record sets you can save as a list, and creates or updates records once you approve | Returns answers, candidate summaries and generated documents such as an offer letter |
| What a search readsThe corpus, not the ranking. | CVs, notes, emails and call transcripts, semantic and keyword together | Global Search across candidates, matches, vacancies, companies, talent pools, leads and opportunities; keyword search on CV terms, skills and education |
| The shortlist | ||
| Why a candidate is on itCan a consultant defend the order to a client? | A pass or gap verdict and the supporting line from the record, for every requirement | Matches proposed automatically and reviewed in a screening overview; per-criterion reasoning is not described publicly |
| Where the criteria come fromWhose rules the ranking runs on. | Criteria you write and edit in plain English, per job | Assisted Matching on the vacancy; deeper AI matching comes from theMatchBox, a partner integration |
| What the ranker can seeA shortlist is only as good as the parsing under it. | Bulk CV import and parsing included, on the engine behind RemakeCV | Automated CV Parsing from the Premium edition up |
| Trusting the record | ||
| Field-level historyWhat changed, who changed it, and from what. | Per-field history with the source of every AI suggestion, and undo | Activity is logged automatically against the file; per-field history and undo are not described publicly |
| Two people, one recordThe quiet data-loss bug in most CRMs. | Optimistic concurrency, so a stale edit is refused rather than silently applied | Not described in their public material |
| What a conversation does to a recordApplied, or proposed for review? | Emails, calls and messages become proposed field updates and tasks you accept or reject, with the source attached | CX Sense prepares a next step you check and send; WhatsApp conversations enrich the file automatically |
| Publishing and plumbing | ||
| Careers siteWhether the roles rank for your brand. | Included: your own domain, SEO-indexed, applicants parsed and scored on arrival | A recruitment website is bought separately or built with a partner; a WordPress plugin publishes vacancies to it |
| WebhooksGetting events out to your own systems. | Included, with delivery logs and redelivery, alongside a scoped public REST API | Rolling out gradually on request through your CSM; access sits in the paid Carerix Datasource bundle |
| Getting your data back outThe question to ask before you sign, not after. | Full workspace export at any time, on request, with no exit fee | Data Connector from the Premium edition up; export terms are not published |
| What you actually buy | ||
| EditionsHow many decisions before you can buy. | One edition. Every feature included, unlimited records and storage | Four editions, each adding features and a larger allowance of “taken” (tasks) |
| Published priceWhether you can compare without a sales call. | USD 119 per user / month billed annually, on the website | Does not publish pricing; every edition is demo-and-quote |
| Usage capsWhat happens when the team uses the thing. | Unlimited records and storage; an AI allowance included, or an included monthly AI allowance | A task allowance per edition, from 1,000 on Starter to 50,000 on the top edition |
| MigrationWho does the unglamorous half. | Done for you on every plan — extraction, mapping, de-duplication, and fifty of your own records to check first | Not published; implementation runs as a project, with user and administrator training through Carerix Academy |
Carerix Private AI is scoped on purpose: their own page states the system is fenced to candidate, vacancy and matching support, and that asking it anything else returns an error. Inside that fence it is genuinely useful — it answers questions about a candidate’s data and writes summaries, interview questions and offer letters. Kepler’s assistant is a sidebar on every page with about fifteen tools: it searches CVs, notes, emails and call transcripts across the entire workspace, returns record sets you can save as a list, and creates or updates records once you approve the change.
Active data engineers in London who have used Kafka in production, available within a month
Found 4 candidates. Three mention Kafka on their CV; one said it on a recorded call last week.
| Candidate | Current employer | Notice | Why it matched |
|---|---|---|---|
| Sophia Bennett | Monarch Analytics | 1 month | Kafka streaming for card transactions at Halcyon CV |
| Amir Hassan | Northwind Capital | 2 weeks | Kafka feed for the risk platform CV |
| James Okafor | Juniper Systems | 3 months | Rebuilt the Kafka clickstream pipeline CV |
| Oliver Chen | Northwind Capital | Immediate | Mentioned running Kafka on the 2 Sep call Call note · 2 Sep |
Carerix proposes matches continuously — the Search & Match assistant watches for new candidates and vacancies and puts the results in a screening overview for a recruiter to judge, and theMatchBox adds a deeper AI matching layer as a partner integration. Kepler’s matching runs on criteria you wrote for that specific job, and every requirement expands under the row into a pass or gap verdict with the line from the candidate’s record that supports it, so the order can be defended rather than trusted.
| Candidate | Match | Role title | Current employer | Skills | |
|---|---|---|---|---|---|
| Sophia Bennett | 92Strong | Senior Data Engineer | Monarch Analytics | Python, Spark, AWS, Terraform, dbt | |
| Amir Hassan | 81Good | Data Engineer | Northwind Capital | Python, Spark, EMR, SQL, Airflow | |
| James Okafor | 74Good | Senior Data Engineer | Juniper Systems | Spark, Azure, Databricks, Kafka, Scala | |
| Mei Tan | 68Weak | Engineering Manager, Data | Atlas Cloud | AWS, Leadership, Python, Snowflake | |
| Elena Rossi | 55Weak | Analytics Engineer | Forma | dbt, SQL, Python, Looker | |
| Oliver Chen | 49Off spec | Platform Engineer | Northwind Capital | AWS, Kubernetes, Terraform, Go |
Assessment
A strong fit for the brief. Sophia has run a production Spark platform on AWS and has managed a team through a migration. Financial services exposure is via a payments client rather than a bank.
Criteria · 3 of 4
Decision
↑ ↓ next candidate · Esc close
Both products mine conversations. Carerix CX Sense reads signals out of calls, notes, emails and workflows and prepares the next step — typically a message you check, adjust and send — while their WhatsApp Conversational AI enriches the file automatically as a candidate replies. Kepler routes every derived change into a proposed field update or task that a person accepts or rejects, with the email, call or message it came from attached, and keeps that source in the field’s history afterwards.
Three candidates · this week
Re: Senior Data Engineer at Northwind
Thanks for sending the details over. The role sounds great, and I can start on 3 October if the timing works for them.
Are you still keeping half an eye out?
I handed my notice in this morning, so I am actively looking now.
James · 11:26For my next move I am looking for £105,000. I would want a step up for a move.
| Candidate | Available from | Status | Expected salary | Notice period | Company | Owner | Skills | |
|---|---|---|---|---|---|---|---|---|
| Sophia Bennett | Not set | Passive | £90,000 | 1 month | Monarch Analytics | Priya Nair | ||
| Amir Hassan | 14 Oct 2026 | Passive | £85,000 | 2 weeks | Northwind Capital | Priya Nair | ||
| James Okafor | Immediate | New | £100,000 | 3 months | Juniper Systems | Tom Reilly | ||
| Mei Tan | Not set | Passive | S$160,000 | 2 months | Atlas Cloud | Tom Reilly | ||
| Elena Rossi | Not set | Active | €78,000 | 1 month | Forma | Tom Reilly | ||
| Oliver Chen | Immediate | New | £88,000 | None | Northwind Capital | Priya Nair | ||
| Rahul Menon | 2 Nov 2026 | Active | £92,000 | 1 month | Juniper Systems | Priya Nair | ||
| Yuki Tanaka | Not set | Passive | ¥14,000,000 | 3 months | Atlas Cloud | Tom Reilly | ||
| Hafiz Rahman | 20 Oct 2026 | Active | S$145,000 | 2 weeks | Meridian Partners | Priya Nair | ||
| Clara Nowak | Not set | New | €84,000 | 1 month | Forma | Tom Reilly | ||
| Daniel Okonkwo | Immediate | Active | £78,000 | None | Northwind Capital | Priya Nair | ||
| Ana Ferreira | 8 Dec 2026 | Passive | €96,000 | 3 months | Juniper Systems | Tom Reilly | ||
| Tomas Novak | Not set | Active | €71,000 | 1 month | Monarch Analytics | Priya Nair | ||
| Grace Adeyemi | 27 Oct 2026 | Active | £105,000 | 2 weeks | Atlas Cloud | Tom Reilly | ||
| Ivan Petrov | Not set | Passive | €88,000 | 2 months | Forma | Priya Nair | ||
| Leila Haddad | Immediate | New | £82,000 | None | Meridian Partners | Tom Reilly | ||
| Marcus Reid | 16 Nov 2026 | Active | £98,000 | 1 month | Monarch Analytics | Priya Nair | ||
| Nadia Kaur | Not set | Passive | S$172,000 | 3 months | Northwind Capital | Tom Reilly |
Kepler's CV parsing and formatting are ported from RemakeCV, our own product, which 300+ recruitment agencies use. The migration that fills your database on day one runs on an engine that has been reading real agency CVs for years, not on something written for this launch.
RemakeCV is a Kepler product. It is the same team and the same code, which is why we are willing to point at it — and it is the only outside evidence on this page, because Kepler itself is new and has no customer stories yet.
Whatever your current system gives you, plus any loose CVs and spreadsheets sitting in folders. If you are not sure how to get the export out, we will walk you through it.
Extraction, mapping, de-duplication and validation — the unglamorous half most vendors leave on your desk. Your custom fields are recreated rather than flattened into notes.
Fifty of your own records, fully parsed — your people, not a demo set. Tell us what we got wrong and we fix it and re-run, as many times as it takes.
The full database goes in, mailboxes and calendars connect, and history backfills onto the right records. We stay on it until your team is working in Kepler.
£89 per user / month, billed annually
Does not publish pricing
Carerix publishes no price. Their Next Editions page lists four packages — Starter, Essential, Premium and a top edition — and describes what each one adds, including an allowance of “taken” (tasks) that runs from 1,000 on Starter to 50,000 on the top edition, but carries no figure and no per-seat rate. Every route on the site ends at “vraag een demo aan”. Checked 6 Sep 2026.
Two things worth knowing before you compare. Automated CV Parsing and the Data Connector start at the Premium edition, and the portals and the mid- and back-office integration sit in the top edition only, so the edition you would actually need is probably not the one at the bottom of the ladder. And the Dutch and English versions of that page do not agree on the top edition — the Dutch page calls it ARF and lists mid- and back-office integration, the English page calls it Enterprise and lists the client, candidate and supplier portals. Ask Carerix which list applies to your quote.
None of that is a criticism. Quoting an editioned platform per customer is normal in Benelux staffing, where the back-office integration is half the implementation. It does mean the number that matters to you is not on their website and cannot be worked out from it. Kepler publishes one number and puts everything inside it.
It depends on whether you bill placements or hours. Carerix is the better product for a Dutch or Belgian detachering, uitzend or ZZP agency: their secondment edition generates contracts, handles timesheet registration and drives invoicing from the same system as the recruiting, their client portal lets a customer approve hours online, they integrate AFAS, Exact, Twinfield, Youforce, Easyflex and HelloFlex, and they hold ISO 27001. Kepler has no timesheets, no invoicing, no payroll and no back office at all. Kepler is the better product for a permanent or contingency agency whose asset is a large database nobody can search — one assistant across CVs, notes, emails and call transcripts, matching with per-requirement evidence, per-field history with undo, and one published price with every feature switched on.
Carerix does not publish pricing. Their Next Editions page names four packages — Starter, Essential, Premium and a top edition — and lists what each adds, including a task allowance rising from 1,000 to 50,000, but shows no price for any of them; every path on the site leads to a demo request. Kepler publishes USD 119 per user per month billed annually, which is USD 1,428 per seat per year, with one edition, every feature included, unlimited records and storage, free migration, and additional AI capacity on request. Checked 6 Sep 2026.
Yes, and more of it than most platforms of its age. Carerix Private AI includes system, user and custom prompts, an AI assistant on vacancies that rewrites and checks job text, an AI assistant on candidates that answers questions about candidate data and generates documents, Conversational AI over WhatsApp for pre-screening and follow-up, and Assisted Matching. CX Sense reads signals out of conversations, notes, emails and workflows and prepares the next step. Their own page states the AI is deliberately fenced to candidate, vacancy and matching support and that general questions return an error, which is a reasonable safety decision and also the main difference from Kepler, whose assistant works across the whole workspace and can write changes back after approval.
Yes. The Carerix Notetaker attaches to your calendar and appointments, records digital meetings and live conversations, produces a summary and files it against the right candidate, contact, vacancy, match or placement — and it is available in their mobile app, including recording an in-person conversation or uploading an audio file afterwards. Kepler’s notetaker joins the call and files the summary, the next steps and the full transcript against the record. This is a capability both products have and it is not a reason to pick one over the other.
No. Kepler has no timesheets, no invoicing, no payroll and no back office of any kind, and this is the single biggest gap between the two products. Carerix describes its detachering edition as a complete system in which you recruit, select and place, generate contracts, handle timesheet registration and drive the invoicing process, with a client portal for online approval of hours and integrations to AFAS, Exact, Twinfield, Youforce, Easyflex and HelloFlex. If your desk bills hours, Carerix is the right software and Kepler is not. Kepler’s answer is the integration story — a public REST API with scoped keys and outbound webhooks with delivery logs — not a module.
Kepler runs the migration for you on every plan rather than handing you an importer. You send an export from Carerix plus any loose CVs, spreadsheets and shared-inbox archives that never made it into the CRM; Kepler handles extraction, mapping, de-duplication and validation, and rebuilds your custom fields as real fields rather than flattening them into notes. Before the full load you review fifty of your own records, fully parsed, and anything wrong is corrected and re-run. Activity, stages and history come across, not just names and email addresses. Kepler also exports the whole workspace on request at any time with no exit fee. The honest caveat: if any part of your Carerix setup is doing contracts, hours or invoicing, that part has nowhere to land in Kepler.
Bring your hardest search to a 15-minute call. We will run it live, on real records, and talk through what a migration would look like.