About Us

Why organisations choose Prospero

Organisations choose Prospero because DetectX® gives banks, insurers and other regulated organisations results they can explain and controls they can configure, backed by a team with more than 25 years in regulated industries. Proudly Swiss made.

Experience in regulated industries
Prospero has worked with banks, insurers and enterprises for more than 25 years, and that experience shapes every DetectX® solution.
All industries
Explainable AI and machine learning
DetectX® applies machine learning in ways your team can explain and audit. Every score shows its reasons, and every action is recorded.
DetectX® Overview
Fewer false positives, less alert fatigue
Thresholds, whitelisting and alert filtering hold back repeat noise, and name matching weighs date of birth and nationality, not spelling alone.
Name Screening
Configured by your team, to your rules
Thresholds, rules, lists and workflows are yours to set and adjust. Start with one solution and add more as your needs grow, on one platform.
All solutions
Integrates with your systems and partners
DetectX® is modular and data-agnostic. It connects to core banking, fraud tools and partner platforms, directly or through a data warehouse.
Partners
A planned, phased implementation
A single screening module is typically a 15‑week project. Broader scopes are delivered in phases, each tested with your team before go-live.
Talk to us

From data to decision

Every DetectX® solution follows the same path. Your data comes in, the analytics engines profile and score it, and whatever needs a person reaches one, with the evidence already attached. What changes between solutions is the data and the rules, not the platform.

Your data comes in

DetectX® works on the data you already hold. Customer records and payments arrive from your core system, directly or through your data warehouse.

In real time: New customers, customer updates and payments are checked the moment they arrive.
Overnight: Batch runs re-examine your full client base and past transactions.
Lists you choose: Sanctions and PEP lists are imported from the providers you select.

Every engine reads it

Each record is read several ways at once. Names are compared with every list, behaviour with the client’s own history, and connections between parties are traced.

More than spelling: Name matching also weighs date of birth and nationality, and returns a score.
Behaviour over time: Statistics such as all outflows over the last 12 months are calculated for every client.
Who is connected: Link analysis groups the alerts that belong together.

Your threshold filters it

Most records pass without anyone touching them. You set the line, and only what crosses it becomes an alert.

Your rules: Similarity thresholds, business rules and risk scores are configured for your institution.
No repeat noise: Alert filtering can hold back further alerts on the same case for a set interval.
Risk sent back: Client risk scores set a risk class that is returned to your core system.

Your team makes the call

An alert is a question, not a verdict. AI recommends a next step and explains why. Your team investigates in AlertViewer and makes the decision.

AI recommends: Each alert gets a recommendation, its reasons and a confidence score.
Decide and act: Approve, block, whitelist, put on re-watch, or report to the regulator.
On the record: Every user action is recorded for transparency and audit.