Waltham Forest is by far the coolest borough in London; it’s located in East London and has a vibrant and diverse community, making it the ideal title-holder for the first ever London Borough of Culture.
Waltham Forest is a predominantly residential borough, with one of the smallest economies in London. High population growth during the past decade has been driven by international migration, and we have a high rate of population churn, with 74 per cent inflows from other London boroughs. We have a young demographic compared with the UK average, with over 65,000 children and young people representing about a quarter of the population and this is a big part of why Waltham Forest is so vibrant and cool.
To put it simply, we are awesome.
We do have our problems. We have a rapidly-aging population and we were ranked as the 35th most deprived local authority in England in 2015. We are under extreme budget pressures and we have to look for better ways to provide public services for less.
We have approximately 2,800 officers and 60 councillors running services for 280,000 residents. We run everything you can imagine, from collecting rubbish and recycling to building control, social care, housing, libraries, parking and roads.
We have an IT department of 70 staff, with over 150 systems, 85 customer portals and websites, and one chatbot called Walt, who allows residents to report fly tipping.
Dispersed data
We have loads of data, but it’s very difficult to aggregate information to produce reliable real-time results. We have data silos, complex legacy architectures and a lack of shared insight across the organisation.
We wanted to bring information together across our systems, understand our customers and gain insights about the services we offer. We also have statutory responsibilities which we need to protect and we want to keep our streets clean and safe for our residents.
We didn’t want to run a data warehouse project to bring information together and spend an enormous amount of time developing it only to realise the benefits in three or four years’ time. We needed to be fast and gain immediate value from our services and, above all, the project had to be ‘smart’.
Ultimately, we needed to resolve a problem, so we picked one and created a proof of concept.
Tackling rogue landlords
Waltham Forest has a ‘selective licence’ scheme for private landlords. The scheme is designed to allow local authorities to tackle rogue landlords who aren’t taking managing their properties responsibly. Our landlords need to adhere to conditions we set in our licences in order to rent out their property. Our problem is that we don’t know if we have all the private landlords identified. Social landlords and home owners are also exempt but we don’t know where they all are in the borough. Lastly, we wanted to deal with antisocial behaviour and improve our communities.
The data is spread across five separate systems, So, working with Amazon Web Services (AWS) and NG Data, our first step was to match the information and create a data lake. Using the power of AWS in the cloud and Informatica technology, we were able to connect and match systems and propagate a data lake. The system was able to intelligently work across the data schemas and provide a workflow to connect to other systems. In addition, it made guesses of how to match it up, cleansed the information and introduced master data management across the entire platform.
After some minor amendments to the workflow engine to verify that the information was matching correctly, we moved on to produce a real-time dashboard to display the information. The data was in an easy-to-use format to allow officers to drill down for further investigation if needed.
However, that’s not the end of the story. We could clearly see the properties that we knew about (i.e. those that were already licensed) and we also had social housing properties. But what about the properties we didn’t know about?
Making predictions
We needed an algorithm which could predict the likelihood that a property could be rogue. By looking at key indicators from the datasets, we could calculate a score to show the likelihood that a property could be run by a private landlord; for example, for any property which has regular changes in council tax over a short period of time as well as changes in electoral role, then there is strong possibility that we have a private landlord. Again, if we look at antisocial behaviour reports and other indicators such as pest control, then we could have private landlord who is not licensed and is not dealing with those problems.
Creating a regression model as the second step and using the power of AWS, we were able to provide indicators for likelihood and data accuracy. This means officers have percentage scorings to indicate if there is a high likelihood of a private landlord and that the data really is clean. At this point, the officer can open an investigation and look for evidence about whether they are keeping to the terms of the licence.
The dashboard also shows information for antisocial behaviour, fixed penalty notices, pest control and legal summons, all of which is useful indicative data to understand how a private landlord is performing.
Why is this important?
The smart data lake in this specific proof shows that we can rapidly connect information and then visualise data. This means that for the first time we can help officers to target the right the properties, rather than manually trying to find out what’s going on. It also means that we can be proactive and stick to our promise to make Waltham Forest clean and safe for our residents.
We also know that we can update information direct to our back-office systems using the smart data lake. This means we can plug in apps, sensors, chatbots, AI and smart devices to create the beginnings of a ‘smart borough’ and then automate services. For example, we can use the data to power intelligent chatbots to provide information to residents without needing to have a face-to-face conversations, so that our officers can then focus on the residents who really need our help.
Information is essential for understanding our residents in ways we couldn’t do before. By plugging in our CRM system, we can gain insights into the needs of our customers and provide better services; for example, if we can identify that we have a university student living in the borough, then perhaps we can send them a welcome pack to help them make best use of our services. In fact, what we would love to do is to help that student stay in the borough and start a business when they finish their studies.
Prevention is better than cure
Another way to think about how bringing data together can help our residents is in terms of prevention and intervention at the right time. If we can understand which families have children who end up in a gang, for example, perhaps we can work with our colleagues in the police and neighbouring councils to prevent our younger population from making bad choices in the first place and offer opportunities which really inspire them.
We want to continue to provide great services to our communities and, considering our very tight financial pressures, the only way to do that is by bringing our data together in order to provide a smarter way of delivering those services.
The smart data lake is a cornerstone for the future platform, and that is awesome.
Richard Holland is the assistant director of technology and innovation at Waltham Forest Council.
