Most AI recruitment shortlisting tools are ATS add-ons.
They sit inside your existing applicant tracking system. They scan CVs. They spit out scores. They promise to save you time.
But they rarely do.
Because they are built for HR departments with compliance budgets. Not for recruiters who need to fill roles fast.
Lucuma does it differently.
We built it as a standalone tool. No ATS integration required. No dependency on clunky enterprise software. Just upload the CVs, paste the job description, and get a ranked shortlist with a reason attached to every score.
Here is why that matters.
Most AI Recruitment Shortlisting Tools Are Buried in ATS Platforms
ATS systems are designed to track candidates. Not to help you make faster hiring decisions.
AI features get bolted on as upsells. They inherit the same slow interface. The same bloated workflows. The same need for admin access and IT approval.
You still have to log in to a platform you hate. Navigate three menus. Wait for scores to process. Export results to share with hiring managers.
Lucuma skips all of that.
It is a web app. You go to lucumashortlist.cv, drop in your CVs, and you are done. No onboarding process. No training sessions. No IT tickets.
Recruiters do not need another tool they have to learn. They need something that works immediately.
We Rank Candidates by Fit, Not Just Keywords
Most AI recruitment shortlisting systems scan for keywords.
They look for "Python" or "project management" or "five years' experience" and assign a score. If the CV has the words, it ranks high. If it does not, it ranks low.
That is not shortlisting. That is Ctrl+F with extra steps.
Lucuma reads the job description and the CV together, not as two lists of words. It weighs context. It recognises transferable skills. Someone who "led cross-functional teams on product delivery" may well fit a role asking for "stakeholder management experience," even if those exact words are not there.
We use LLMs to analyse tone, structure, and relevance. Not just surface-level matching.
The result is a ranking based on what the CV actually says, not on which words happen to appear in it.
No Black Box Scoring
One of the biggest complaints recruiters have about AI recruitment shortlisting is the lack of transparency.
A candidate gets a score of 78. Why 78? What did the AI see? What did it miss?
You cannot explain it to a hiring manager. You cannot justify it to a candidate. You just have to trust the algorithm.
Lucuma shows you the reasoning.
For each candidate, you get a summary of why they ranked where they did. What made them a strong match. What gaps the AI spotted. What stood out in their CV.
Nothing is hidden behind the number.
You can take that summary into a conversation with a hiring manager and explain exactly why you are putting someone forward. Or why you are not.
Transparency builds trust. And trust makes AI tools actually usable.
Built for Agency Recruiters and In-House Teams
Most ATS platforms are built for big companies. They assume you have an HR team. A compliance officer. A budget for enterprise software.
Lucuma is built for recruiters who work fast.
Agency recruiters handling 10 roles at once. In-house talent teams at startups with no ATS budget. Freelance recruiters who need a tool that works on their laptop without a login wall.
You do not need to integrate it with your email. You do not need to map custom fields. You do not need to assign user permissions.
You just need to shortlist candidates quickly and accurately.
That is what we optimised for.
Explainable by Default, Because the Rules Now Require It
UK rules on automated decisions moved in 2025. Under the Data (Use and Access) Act, a significant decision made solely by software, such as rejecting an applicant, needs meaningful human involvement, notice, and an explanation the person can contest.
"The software decided" is not a defence.
Lucuma never rejects anyone on its own. It scores and ranks CVs against the job description. A person decides what happens next.
Every score carries a plain-English reason, so when a candidate asks why, you have an answer on the record instead of a number you cannot account for.
It scores against the criteria you set for the role, not against a model's guess at who worked out before. Name, age and gender are not factors. None of this is legal advice and it does not move your responsibility onto us, but it is a process you can defend.
Speed Matters, But Not at the Cost of Knowing Why
Speed matters in recruitment.
A great candidate applies on Monday. By Wednesday, they have three other interviews lined up. If you have not shortlisted them by Tuesday morning, you have already lost them.
AI recruitment shortlisting should make that process faster. Not slower.
So Lucuma takes the CVs for a role in one go, scores each one against the job description, and hands back a ranked list with the reasoning attached.
No waiting for batch processing. No overnight syncs. No "results will be ready in 24 hours."
Just a ranked list you can explain.
Why We Built Lucuma This Way
At Marvanova, we build tools for founders and small teams who need AI to work practically.
We do not build enterprise software. We do not build tools that require six-month implementations. We do not build features that look good in a pitch deck but frustrate users in real life.
Lucuma came out of the same frustration recruiters have with their ATS. Scores you cannot question. Workflows you cannot skip. A tool you have to ask IT for permission to open.
So we built it standalone. No integrations. No dependencies. A transparent AI recruitment shortlisting tool that does one thing: score and rank CVs against your job description, and show its working.
If you are tired of ATS add-ons that promise the world and deliver keyword matching, try Lucuma.
Go to lucumashortlist.cv and run it on a role you are hiring for right now. See what it says, and why.
Or email us at hello@marvanova.com if you want to talk about what we built and why.