RHOSTONE
Services

We model the system behind the question.

Most of the questions we take on cross finance and law, and the facts sit in data, contracts and rules held by different teams. We bring experience in credit risk modelling, machine learning, technology law and software, test the model on your own data, then advise you or build the software that answers it.

What we offer

Two kinds of advice, and software when the answer needs it.

Tested answers and legal tech advisory, either of which can lead to a buildFOR FINANCE AND LAW QUESTIONSANSWERTested answersLegal and financial risk modelled onyour own data, with a model you canrerun.FOR LEGAL TECHNOLOGY COMPANIESADVISELegal tech advisoryContracts and rules turned into datasoftware can use, and which legal workto automate.WHEN IT NEEDS SOFTWAREBuildDecision tools for workthat has always neededhours of humananalysis.Tested answers and legal tech advisory, either of which can lead to a buildFOR FINANCE AND LAW QUESTIONSANSWERTested answersLegal and financial risk modelledon your own data, with a modelyou can rerun.FOR LEGAL TECHNOLOGY COMPANIESADVISELegal tech advisoryContracts and rules turned intodata software can use, and whichlegal work to automate.WHEN IT NEEDS SOFTWAREBuildDecision tools for work that hasalways needed hours of humananalysis.
Two ways in. Either can end in a build
How we work

Five stages, taken on their own or together.

Five stages, what each leaves behind, the loop from test back to model, and which engagement covers each01UnderstandObserveWe map thesystem: itsdata, documentsand rules.02ModelHypothesisFinance and lawin one model,with forecasts.03SimulateExperimentScenarios andstress testsshow where theplan breaks.04EncodeRuleEach rulebecomes codethat cites itssource.05BuildApplicationA working toolyour team ownsand runs.RESULTS GO BACK INTO THE MODELDIAGNOSEPROTOTYPEBUILDFive stages, what each leaves behind, the loop from test back to model, and which engagement covers each01UnderstandObserveWe map the system: its data,documents and rules.02ModelHypothesisFinance and law in one model,with forecasts.03SimulateExperimentScenarios and stress tests showwhere the plan breaks.04EncodeRuleEach rule becomes code thatcites its source.05BuildApplicationA working tool your team ownsand runs.RESULTS GO BACK INTO THE MODELDIAGNOSEPROTOTYPEBUILD
Few problems need all five

We agree scope, timing, data access, confidentiality and intellectual property with you in writing before work starts.

Questions

Before you get in touch.

The problem is still vague. Does that matter?

No. That’s what Diagnose is for.

How do you handle confidential data?

We agree access, security, retention and which AI tools may be used before you share anything sensitive. Please keep your first email high-level.

Is this legal advice?

No. RhoStone isn’t a law firm and doesn’t give legal advice. Where you need it, we work alongside your own counsel.

Not sure where to start?

Describe the problem and we’ll agree the first step together.

Describe the problem →