Cifas Case Study
Blogs home Featured Image

Case Study: Helping Cifas further its intelligence capabilities

Cifas intelligence is used to identify fraud activity, organised crime networks and emerging fraud types, and is shared with law enforcement agencies to help detect, prevent and tackle fraud.

By using Mango Solutions’ data science capabilities and expertise, the UK’s leading fraud prevention service is able to harness this intelligence more effectively and efficiently.

Learn how Mango Solutions enables the UK’s leading fraud prevention service to deepen its understanding of financial crime.

 

Mazda
Blogs home Featured Image

Data is integral to the DNA of Mazda’s business, with the analytics resource being used to drive more effective decision-making in terms of financing, forecasting market needs of all sorts, and in-depth customer understanding.

Read more about how Mango worked with Mazda to enhance their internal analytic capability, enabling them to move towards the next phase of the company’s customer data-driven journey.

pure planet
Blogs home Featured Image

Britain’s first all-digital, renewable energy supplier Pure Planet harnesses data to drive service efficiency

Pure Planet was founded by friends who wanted to offer green energy for less cost than brown, polluting power – why would people pay more to pollute, when they could pay less to be clean and green?  

Its mission is to enable a renewable powered Britain by making renewable energy mainstream – and price competitive – while delivering technologicallyadvanced services that are simple to use and easy to access. The first energy provider to offer 100% renewable electricity and 100% carbon offset gas, it is certainly shaking up the industry, offering a price sensitive, adaptable and clean choice.  

This young, independent, Bath-based dynamic company was launched in 2017 to drive the rapid adoption of clean, renewable energy and now employs more than 100 people in the UNESCO world heritage city. It’s multi award-winning, and is the Best Company in the South West and the Second Best Smaller Company to work for anywhere in the UK, according to The Sunday Times/Best Companies. 

The business is built around the application of smart, innovative technology for the dedicated requirements of its expanding membership base, making sustainability an easy choice. 

Challenge

Pure Planet faced a challenge in the Data Science area after its first year or operation. With the team lead on sabbatical, they wanted a Consultant who could simply step in, get to grips with delivering and harnessing value from their data with immediate effect. With an abundance of data from their chat bot, known as ‘WattBot’ and a variety of other sourcesbeing data-driven has always been central to their strategy and an integral part of their business ethos – helping to drive data-driven pricing solutions, drive service efficiency and to give their members a seamless experience. 

Pure Planet’s Data team of four were using Machine Learning and deploying advanced analytics in many areas including market forecasting evaluation and membership acquisition and churn. It was clear that value could be added in terms of broadening the scope and skills of the data science team and in helping them to establish common frameworks and processes to make data science easier with repeatable and scalable models. 

Solution

In deploying Mango’s expertise, our Consultant was immediately able to demonstrate quantifiable value to the existing team and organisationIn conducting an audit of current practices, Mango was able to identify process improvementsassist in developing a robust data infrastructure and deployment aligned to business objectives   

Mango was also able to assist in the development of an updated Management dashboard, which meant that stakeholders from across the business could have immediate access to the data that was important to them.  This centralised dashboard empowers teams across Pure Planet with the necessary information to make informed data-driven decisionsThe volume of accessible data is now ten times larger than it was, making decisions even more effective, ensuring that effective modelling strategies can be put into place. The team can now assess and predict the likelihood of a Member switching suppliers through an assessment of account their health, payments and meter readings, alongside effective marketing strategies that reduce customer churn. 

Benefits

Pure Planet’s engagement with Mango Solutions represents the perfect partnership approach to data science consultancy. By embedding a Consultant as an integral part of the team, Pure Planet was able to enhance its data-driven solutions that drive competitive advantage and industry innovation.   

The infrastructure upgrades will handle the increased demands and scale required by this rapidlygrowing, innovative business. Pure Planet is now on a course towards more democratised data and a more self-service approach; supported with a machine learning production system.  

By effectively integrating data from various systems, Pure Planet now has a fuller picture of each Member’s experience. The data integrity processes, now in place, help to identify problems and ensures that Pure Planet can make sure each Member’s experience is as smooth as possible. By focusing on building reusable data pipelines for reporting and machine learning, it has enabled processes to be more efficient, with less time spent on building similar reports and more time on building Pure Planet’s advanced analytics capability.   

 “Working with Mango has been rewarding,” said Andrew Ralston, CEO, Pure Planet. “Mango stepped in to provide people cover and has added much more, helping to develop our data science strategy and capabilitiesThe Consultant Mango placed with us has been an excellent addition to our team, and were alo delighted to have been able to draw on the wider support and expertise of Mango. 

Related stories: Graduate Data Science Placement at Pure Planet

C-path tuberculosis drug trials
Blogs home Featured Image

Shiny development enables analysis to support development of treatment regimens for Tuberculosis.

Critical Path Institute (C-Path) is a non-profit, public-private partnership whose mission is to catalyse the development of new approaches that advance medical innovation and regulatory science, accelerating the path to a healthier world.

Orchestrating the development of ‘drug development tools’ (DDT’s) for life threatening diseases, C-Path aim to reduce the overall cost of medical product development through their innovative approach to the sharing of data and expertise.

Tuberculosis currently affects a third of the world’s population and kills more than 1.7m people annually. Through C-Path’s TB Data Collaboration Platform, researchers can have access to high quality, curated and pooled clinical trials patient data that can be used to accelerate TB research and the development of safer and more effective therapies.

Challenge

C-Path wanted to enhance the overall data utility and improve the user experience of the TB Data Collaboration Platform – providing researchers with the ability to easily interrogate and visualize the pooled clinical trials data in the platform, and quickly find the information needed to support their research goals.

Having worked with Mango on previous projects, C-path were confident Mango had the desired skills and technical capability to meet their requirements – to develop more in-depth data exploration and visualization tools, giving researchers the ability to understand and interrogate the data more effectively.

Solution

The development of a front-end R Shiny exploratory interface solution, met C-Path’s functional requirements and provided them with a powerful, flexible and extensible tool that could be further customized to meet the changing needs of the platform and utilized in future data collaboration projects. The solution met the needs of the project sponsor and stakeholders and significantly enhanced the existing functionality of the platform.

Through an agile development process consisting of multiple sprints, Mango worked closely with C-Path to ensure the R Shiny application was developed in line with their schedule and budget. Mango’s agile approach meant that the development could reactive quickly to changes in requirements and scope.

The development met all the requirements and more, said Vicki Theurer Crider, Project Manager of C-Path’s Data Collaboration Center.

“We have really enjoyed working in partnership with Mango and their collaborative working style, expertise and value. Mango excels at working in an agile way, adapting to changing circumstances as they arise. Their communication, dedication and expertise in both a technical and project capacity are exemplary. We have forged fantastic relationships across the team and the benefits of our collaboration are already being realised”.

Results

Having set out to improve the knowledge discovery process for TB researchers using their TB Data Collaboration Platform, C-Path have achieved their aims. The Mango-developed interrogation and visualization application, provides improved accessibility of the data, allowing for more useful and efficient research to be undertaken by developers and researchers working on better TB treatments.

 

Data Science: Boosting Customer Retention in leisure industry
Blogs home Featured Image

About our  client

The leisure company’s success is built around a flexible membership model. With that flexibility comes the challenge that members can cancel their membership without warning. In these circumstances, boosting customer retention is a critical driver of success. The company wanted a data science partner that would focus on delivering value from investments and to help it to develop a stronger analytical culture internally. As specialists in delivering value through data-driven decision making, Mango Solutions partnered with the leisure company to help it create descriptive, diagnostic and predictive insights that have delivered measurable value with a significant return on investment (ROI).

Challenge

In a highly competitive marketplace, the leisure company were interested to find out which of their members were most likely to cancel their membership in the near future. This awareness would enable the business to create tailored marketing campaigns that target the right members at the right time and support retention efforts. This strategy would ensure healthier and more stable membership numbers, which would contribute to the firm’s continuing success and growth.

While some membership loss is inevitable in a highly competitive market, it was believed that retention performance could be further improved. The company recognised that a data-driven solution was a key enabler to successfully overcome this challenge.

However, the company didn’t have its own internal data science capability and developing a strategic partnership with a consultancy that could bring external data science expertise therefore became essential. The company wanted its partner to work closely with its internal data and analytics team to deliver successful projects with measurable ROI.

Solution

In line with the leisure company’s requirement to build internal capability, an experienced team of data science consultants with strong technical and analytic skills, alongside Rich Pugh, Mango’s Chief Data Scientist, provided strategic advice to the company including the Executive Team. Mango consultants worked in partnership with the company’s data and analytics team, but also spent significant time working across areas of the business to understand day-to-day operations and decision-making processes.

In partnership with the client, Mango used a range of machine learning and other data science techniques to build a range of statistical models with measurable and provable ROI. Mango used its proprietary framework to review the requirements put forward by the leisure company and proposed a plan to deliver them. This included an assessment of the quality of the company’s data, its internal data expertise and technological infrastructure, as well as a plan to cover the proposed solution from data collection and preparation through to predictive modelling and monitoring model performance. The effectiveness and performance of models was evaluated against historical data. Mango continued to iterate and improve the modelling efforts in response to new data and by adding enhancements such as feature engineering.

Benefits

The project has had a positive impact by helping the leisure company to understand member behaviours and to establish the right capabilities to deliver data-led change. The company is now able to target the members with tailored marketing campaigns boosting customer retention, which has delivered significant business value through data science, with a proven and measurable ROI. Mango continues to work with the company and this relationship is helping the company to deliver value from data across a variety of business areas, including the potential impact of alternative models and associated marketing interventions.

Working with the Mango team has been an exciting and immersive experience. The quality of the data science team members is very high, and the working style has been one of partnership. Early metrics show that the initial project has delivered ROI in excess of 100% and we will continue to work strategically to agree further areas of collaboration” – Company Head of Data & Artificial Intelligence

 

12) harnesses data to detect network faults earlier
Blogs home Featured Image

i2O: An evolution from pressure management to intelligent network solutions.

i2O provides a suite of smart network solutions to help water utilities respond to the challenges created by an increasing and urbanising population, extreme weather and ageing infrastructure. Today it has more than 100 utilities in over 40 countries around the world that have installed its solutions. Headquartered in Southampton, UK i2O also has offices in Malaysia, and Colombia.

Intelligent networks make the most of existing infrastructure: reducing leakage and bursts, energy use and CO2 emissions, and operating costs whilst improving service to customers.

Challenge

i2O was a company with market-leading control and monitoring devices to help reduce water leakage.

But water companies needed to do much more to improve network performance, relying extensively on customer feedback to identify problems; on traditional time/risk-based approaches to asset maintenance; and on sporadic projects to improve network performance.

Serious water leaks can still be a huge problem, with a recent water leak in the Finsbury Park area of London affecting a 500 square meter space, damaging many properties, and leaving many residents without water for over 24 hours.

i2O recognised that it was collecting masses of data through its devices, that water companies were collecting other network-relevant data; but these weren’t being harnessed to detect network faults earlier, schedule maintenance on the basis of asset condition, or identify opportunities to optimise network performance.

Solution

Rather than building its own data analytics team, i2O came to Mango Solutions for their extensive experience and skills.

The initial scope of work was to implement a tool to infer the condition of a critical network asset – the PRV (pressure reducing valve). This has enabled water companies to shift to condition based maintenance.  This can significantly reduce the cost of scheduled maintenance as most of this is unnecessary and incurs risk when technicians carry out the maintenance on the network.

This tool has subsequently been included as a feature in i2O’s iNet software tool, a hardware agnostic solution offered in addition to i2O’s logging and advanced pressure management solutions.  It has leapfrogged i2O into the big data AI world of high value analytics.

With just 1,500 devices monitoring the network, iNet does the work of 100 analysts, reviewing data at whatever time of day it arrives.  It brings together network-relevant data and delivers insight from it.  It provides dashboard, table, map, graph and alarm features to make the insight actionable.

Results

i2O’s work with Mango Solutions represents the perfect blend of existing standout technology joining forces with an exemplary data-led strategy.  More than a dozen i2O clients are already using iNet within a year of its launch.  One had reduced leakage by 15% in the north of their city, population 1.2m, within 2 weeks of using the solution based on the insight it provided.

Joel Hagan, i2O’s CEO, commented: “Many companies find it difficult to build and maintain their own dedicated analytics team and even if they do they lack the breadth and depth of experience that a specialist can provide.  Mango Solutions has enabled i2O to accelerate its progress towards its goal of providing the market-leading network monitoring software solution to complement our other market-leading intelligent network solutions for the water industry.”

If you have any queries regarding any of the content held within this case study, please contact Leon Walker.

catastrophe modelling system
Blogs home Featured Image

The insurance industry is ripe for disruption and new fintech start-ups are eager to challenge established players. Mango Solutions worked with an insurance company that was embarking on a major transformation programme. The insurance company relies on sophisticated modelling to help assess the cover it provides to clients, yet its legacy scenario-planning system wasn’t fit for purpose. Mango helped the firm to develop a user-friendly catastrophe-modelling system. This system has helped to boost analyst productivity and allowed the business to take a much more accurate view on potential risk.

Challenge

Major incidents – such as freak weather events – can have a major impact on how an insurance company deals with risks and sets premiums. Insurance firms undertake catastrophe modelling, creating sophisticated models to assess the potential impact of incidents – a key component of a successful insurance business. The creation of these models relies on data-science expertise and high-performance computing.

Until recently, the organisation didn’t have a user-friendly way of creating and testing these catastrophe models as part of its scenario planning. This planning allows the company to assess the validity of its models in different circumstances and to consider key factors, such as the potential risk and financial exposure. The firm wanted to completely rebuild its modelling platform, and as part of this process, make it easier for its analysts to add data from other sources to help improve the quality of scenario planning.

The insurance company aimed to create a new, user-friendly front-end system. After being quoted a large sum for the development by a vendor, one of the insurance company’s internal employees, who was an expert in RStudio Shiny – a package that makes it easier to build interactive web apps from the R programming language – developed his own front-end system. This prototype front-end system was created in just four days.

The senior management team was impressed with this system, but were keen to create a high-quality, production version of the application – and that’s where Mango Solutions stepped in to provide expert advice.

Solution

The insurance company wanted to create the production version of the front-end system by Christmas 2018. Mango allocated two specialists to the project and helped the firm meet a short development deadline, delivering the system to the business in less than 100 days.

The final solution has made it much easier for analysts to bring data into the system, so that complex catastrophe modelling can take place. The web-based front end of the new catastrophe-modelling system includes a graphical user interface, with pop-down menus, making it simpler for analysts to build models and test scenarios.

Whereas the legacy planning system was so complicated that only a handful of users could use the system, is new, user-friendly platform supports over 100 analysts – increasing and enabling the firm to work on many more scenarios than was possible before, providing the firm with a more accurate view on the potential risk it is taking on behalf of its clients.

The success of this project has prompted the company to call on Mango’s services again to look at other areas of its data and development environment – applying similar rigour to other projects in areas of modelling and scenario-planning across the insurance industry, and also ongoing support for a move to the cloud.

Benefits

The senior management team has been extremely pleased with the final result of the work on the new catastrophe-modelling system – both in terms of Mango’s ability to work to a tight deadline. Mango continues to work with the insurance company, refining the product to meet new business requirements. Current estimates suggest that the project has aided in boosting revenues relating to the sophisticated modelling systems at the firm, by up to four times.

safeguarding government data in a centralised environment
Blogs home Featured Image

This central government department is reliant on the work of specialist analysts, such as mathematicians and statisticians. These analysts understand the complex relationships between socio-economic variables and build models for the organisation. The analysts test these models against data, generate results, and then turn these results into a graphical form that is useful to senior people in the government department. While this process has always produced useful insight for the government department, the organisation was keen to move away from a legacy method of working that meant its analysts stored data locally on their laptops. It wanted to build a centralised system and secure environment to enable secure and collaborative data modelling – which is where Mango Solutions provided expert assistance.

 

Challenge

Many of the government department’s analysts were analysing data using “R”, a programming language for statistical computing and graphics, on their laptops. While the process worked, it was also risky – if the analyst mislaid their laptop, then the valuable piece of analysis they were working on could potentially fall into the wrong hands, putting the security of the department’s information at risk.

The department was keen to build a more effective IT infrastructure, but it also wanted the new environment to help its analysts become even more productive. The organisation was looking for a productionisation process that would allow its analysts to work in the same way, but via a much more structured and less risky IT environment. The department didn’t want to put too much restriction on how its analysts work, but they also wanted to make sure that the data modelling work they had undertaken was well-managed and secure.

While that tighter form of management would provide clear benefits to the department, it was also hoped that it would create plus-points for the analysts – rather than data being stored locally, the department was looking for a system that would store the analysts’ work centrally. This centralisation process would aim to feed all data into a single platform, making it easier for analysts to share information and collaborate on models.

It was hoped the centralisation process would also give analysts access to more computing power. Under the existing arrangement, analysts that worked locally on their laptops were restricted by the power of the device they were using. By creating a collaborative network, the department hoped analysts would be able to use the power of a bigger, centralised system to undertake high-level number crunching.

 

Solution

The government department was aware of Mango Solutions’ work with other public organisations, including the Office for National Statistics (ONS), and explained how its analysts were using R on their laptops. They explained how they were looking for a centralised computer system that would allow their analysts to work in a similar way, but to do so across a secure system that would make collaboration and information management much easier.

One potential solution might have been to build the new platform on the cloud. But due to governance issues, and restrictions on how public organisations can use the cloud to store data, the government department was eager to create a system that could be managed on-premise in its own data centre.

Mango listened to the requirements of the department and helped establish an internally managed, centralised infrastructure for its R products. Mango also helped to provision a programmers’ workbench, providing a platform for the analysts to write code.

The government department was keen to make sure the analysts’ data output could be accessed and consumed by other people around the organisation as easily as possible. The publishing platform RStudio Connect was implemented to help the government department’s decision makers access and use the analysts’ data models and graphical outputs.

The centralised system that Mango helped to create for the department is run in a server environment, which means the analysts can still work on their laptops but can also draw on the power of a centralised computing platform. This environment provides the analysts with the capacity they need to both store data and run higher-level modelling.

 

Benefits

The infrastructure that Mango helped to create provides a high-performance platform that allows the government department’s analysts to collaborate in a safe and secure environment. The data-modelling outputs produced by the analysts are now stored in a centralised environment which is used by senior managers to help them make better decisions. Mango continues to help manage the platform and is on hand to provide expert advice as and when required.

Generating insight from financial data by creating a cloud-based database
Blogs home Featured Image

Mango was approached by a hedge fund business that was reliant on a rudimentary way of working, involving an over-reliance on spreadsheets that were held in unconnected silos across the business. By using the expertise held within Mango Solutions’ development teams, this organisation has now stopped its reliance on spreadsheets and instead turned to a cloud-based database environment that provides a modern platform for sophisticated modelling – an easier platform to use on a daily basis and maintain in the long term.

 

Challenge

Data is crucially important to hedge fund firms. By generating as much insight as possible from the information it held, the organisation knew it would be able to discover potential investments to help steal a competitive edge over its rivals and generate better returns for its clients.

Unfortunately, the large majority of the company’s data was held in unconnected spreadsheets which created a number of issues. Firstly, the hedge fund firm wasn’t able to undertake higher-level analysis of the data it held. Second, the spreadsheets tended to be held in organisational silos, making it difficult for employees to share data and work across the business. Finally, it wasn’t possible for the firm to enrich the data it held by bringing in data and expert knowledge from alternative financial sources, such as Bloomberg, Morningstar and S&P Capital IQ, and use this combined information to help its analysts to identify new investment trends.

As the firm’s data was held in spreadsheets, it was difficult for employees within the company to undertake system support and maintenance. This lack of standardisation made it difficult for them to combine its financial information and to increase the scale of analysis. In short, an over-reliance on spreadsheets across the hedge fund firm had become an inhibitor to further business growth.

The company was keen to move away from this legacy mode of working – it would be much easier for its analysts to run their models if it could ingest all of its data into a single database platform. While the main driver of this work was to help its employees to undertake better investment analysis, the secondary motivation was to create a cloud-based system that could be more easily maintained going forwards.

 

Solution

Mango Solutions met the hedge fund business and analysed what the firm was trying to do. This period of discovery allowed Mango to become immersed in the company’s activities and their use of technology. Mango then built a series of proof of concepts to show how it would be possible to create a platform that could ingest data from third-party sources and to combine that with internal sources in a single database environment.

These proof of concepts proved the value of having an integrated data layer, resulting in Mango building a platform to production-level quality. With it now possible to use the database on a day-to-day basis, Mango helped the firm create models that would run on top of this platform. This included the kind of heavy-duty statistical modelling that would help the firm’s analysts to identify new investment trends – it was these opportunities that allowed the firm to steal a competitive edge over its rivals.

Mango built the bespoke system using a range of standard technologies and tools – the Google Cloud platform and the Python programming language. These components were brought together and, with Mango’s expertise in modelling, were used to create a single, cloud-based database platform to meet the demands of the firm’s analysts.

 

Benefits

The organisation is now able to undertake comprehensive analysis, using sophisticated models. In comparison to its pre-existing spreadsheet-based approach, the new cloud-based system gives the firm a database environment that is both secure and easy to maintain. Mango continues to provide support to the hedge fund firm and will help to maintain the system in the future. The combination of better modelling and a more reliable system provides the firm with a strong technological platform, for long-term business growth.

 

Building analytics capabilities
Blogs home Featured Image

Mango was tasked with building unified analytics capabilities across the 2,500-strong financial crime unit of a multinational Financial Services Company in order to help the institution tackle fundamental challenges such as money laundering.

Challenge

Financial crime is a fast-growing concern for major financial institutions. Aware of the growing importance of regulation in the finance sector, the company was eager to implement processes that could help it deal with regulation and reduce potential risks from crime, knowing it could help tackle some of these key concerns by ensuring it had a unified data skills capability around the globe to tackle these challenges.

The company needed to support the data analytics skills development of its fast-growing financial crime unit which had expanded to include individuals in various roles and departments around the world. These individuals held a range of capabilities, from the investigation and identification of potential crimes, through to high-level data science capabilities.

The financial services company wanted to create a framework to provide a common set of analytical skills for each job function in the financial crime unit. This framework would ensure consistency and standardisation across all levels of financial crime analysis around the globe and whilst providing support to individuals who were looking to develop their careers.

Solution

Mango had already worked with the client on an earlier worldwide project to help them boost its analytical capability. The success of this earlier project meant the client’s executive team knew it could trust Mango to develop a high-quality, bespoke training programme for its financial crime unit. They also knew that it was crucial to create a common language around analytics, to enable workers to communicate effectively and improve their capabilities.

A six-month project was established in an attempt to upskill the analytics talent by building analytics capabilities across the Financial Crime Unit unit. Mango started by building and delivering an “Art of the Possible” webinar to build awareness and enthusiasm around the benefits of data particularly in the area of Financial Crime. Mango also developed a “Common Lexicon” initiative. This project created common definitions for data terms such as “optimise” and “model”. It ensured technical and non-technical people could communicate, collaborate and make valuable contributions.

Mango then used its Data Science Radar to assess the capability of the workforce. This tool used an online questionnaire to analyse the analytical abilities of employees. Mango used this information to build a skills matrix, which measured and mapped out the analytical skills of the organisation from basic “level one” to advanced “level four”.

With a strong awareness of analytical capability within the Unit, it was then possible to start focusing on how individuals could improve their skills. Mango created learning pathways which explained the skills an employee would need to progress to the next level and then created a portfolio of training courses that were delivered as webinars.

Results

Mango’s bespoke training programme boosted data analytics capability by firstly helping them to understand the benefits of improved analytical capability. It created a common language and standardised definitions of roles with the financial crime unit globally which meant that every person in the unit knew the skills they possessed and the capabilities they would need to hone in order to move up to the next level as part of a career development pathway.