Data is meant to become insights so lets leverage technological solutions for your actuarial problems and make it happen!

JJPR Advisory

About Me

Joe Jordan

Consultant


Oxford graduate with over ten years experience working internationally, degrees in Maths and Statistics, top grade at STEP and A*A*A* at A-Level.

“I focus on applying my knowledge and problem-solving skills to develop commercial solutions to problems. During my studies I developed an eye for detail and rigour. Throughout my career I’ve become accustom to managing and overcoming both the complexity and ambiguity present in performing one-off projects in order to deliver results.”

Services


JJPR Advisory is a consultancy providing Actuarial, Analytics and Strategy services.

We help companies take the next step in their analytical journey to bring real data driven business value.

Data is meant to become insights so lets leverage technological solutions for your actuarial problems and make it happen!

Products


PFFT: Pricing Framework in Fast Time


PFFT is a framework for developing high performing regression models for insurance pricing in R. If you are interested in adopting the latest statistical and machine learning methods from the R language then you should consider migrating your analytics technology to R. In doing so, PFFT makes the move away from EMBLEM seamless. Even if your not wanting to adopt R as your main language, PFFT is still a great choice as it is compatible with more advanced statistical regression models.

Contact



Paris

What People Say

"Joe has been an asset to the department and his valuable expertise will be missed."

Chief Risk Officer, Major Global Insurer.

28/12/2020

Writing queries in SQL has been a skill I've often seen in short supply. Not only the process of writing the technical syntax, but also the DWH-specific knowledge for the given organisation. By refining GPT-3 with a series of training examples encapsulating the relationship between Natural English Language, the syntax of SQL and the DWH structure, it can transform the way we interact with data. It looks like NLP solutions to this problem feel a lot more realistic and the incredible thing is that GPT-3 is able to learn fast - apparently 10 examples can be enough for some uses. To get an idea of what I'm talking about, see below:

Copyright © 2020 Joe, Continuously Deployed with GitHub and Netlify - Design: Tooplate