Senior Data Scientist in UK

Data Science Role Profile

About Us

Envisso is a global fintech founded by a team with deep financial services experience and with backing from some of the world’s leading Venture Capital investors.

We are reinventing the way credit risk is managed in payments, by unlocking the value of the massive amount of data payments companies have access to.

Merchant Acquiring companies (also referred to as ‘Payment Services Providers’ (PSPs)) face credit risk from the merchants they process payments for, due to the fact that card scheme rules (Visa, Mastercard, Amex and others) require that if a merchant defaults on the obligation to pay money back to a customer through a chargeback, the payments company must step in and fund the payment to the customer.

Depending on the size and type of the merchant this risk can be significant and PSP’s must take measures to protect against the cost of default.

The current industry practice is for PSP’s to take collateral from a merchant, usually through a ‘rolling reserve’ which is where a portion of sales are held back and only paid to the merchant days, weeks or sometimes even months after the original transaction.

Envisso aims to address two major problems with the way credit risk is managed in payments.

Firstly, the approach to measuring credit risk is manual and often based on generic assumptions.

Secondly, using collateral as a risk mitigation tool causes friction with merchants who do not want working capital tied up and ultimately it also provides poor risk protection for the PSP.

Envisso will solve this via the Envisso Platform:

• Envisso Monitoring: uses payments data alongside data from 3rd parties (such as credit reference agencies) to automatically monitor merchants for changes in their risk profile. 

• Envisso Protect: a first of its kind insurance protection against the cost of default by merchants on their chargeback obligations. Premiums are based on individual merchant risk and are charged as a percentage of payment volumes. 

Underpinning these products are four core types of models:

• Anomaly detection - highlighting where payments data suggests a statistically significant change in behaviour has occurred,
• AI (large language) models which ingest data from the merchants website and other public sources (along with their payments data) which identify areas of increased merchant risk,
• Credit models (mostly machine learning based) predicting the likelihood of a merchant ‘failing’, and
• Exposure models predicting the severity of a claim if a merchant does fail.

Within Envisso there is the ability to build and deploy best-practice approaches, but there are also huge research and development opportunities.

For example payments data is a largely untapped, but hugely rich source of information for predicting the merchants credit risk.

Anomaly detection in an industry where seasonality and customer behavioural changes cause lots of ‘noise’ is uniquely challenging.

And predicting the level of exposure (the value of chargebacks we’d see if the merchant failed) is a problem which we don’t believe has been tackled in a robust and comprehensive way before.

We are looking for data scientists to both build using best-in-class techniques, but to also advance the knowledge and approaches in these areas and to help us to continue to build our unique and valuable intellectual property.

We are a remote first organisation so you can be based anywhere globally.

We believe diverse teams win and welcome applications from people of all backgrounds.

Envisso has built a collaborative, flexible and supportive culture where everyone can thrive. 

The role:

We’re looking for someone who is excited at the prospect of facing new and difficult data challenges.

Someone who is comfortable working as part of a team who are building things from scratch.

We’re looking for someone with deep technical skills, but who is really motivated by making an impact on our business and our customers. 

Responsibilities:

• Build, implement and monitor predictive models which help us accurately understand the level of credit risk of the merchants our partners are serving/insuring,
• Develop creative solutions and methods to better model risk by leveraging a variety of different data sources and techniques, 
• Liaise with various payments firms and insurers to understand their needs and help shape the solutions we bring to market,
• Help guide coach and develop the colleagues around you in data science tools and techniques, how to get the best out of your models, and how they can support the data science function (e.g. by helping define the data environment we need to develop)

We are looking for:

• 3-5+ (could be much more) years experience building statistical models in Python and extracting/manipulating data in SQL is essential, 
• 3-5+ years experience in consulting and/or start-up environments related to the financial services industry,
• Credit risk modelling experience including building probability of default models is required (doing so for SME/commercial lending preferred but not essential)
• LLM experience would be beneficial but is not essential, especially if it includes experience of Retrieval Augmented Generation techniques.
• Commercial acumen and an understanding of the economics underpinning credit portfolios would be highly beneficial,
• Able to work autonomously and to display strong judgement and proactivity is essential

What you'll get in return:

• Flexible, remote-first working arrangements
• Opportunities to grow and progress and to work on problems no-one has worked on before
• Competitive package including company equity (anticipated base salary £80k-£100k depending on experience but we're open to applications outside of that range).