Liverpool City Council

Liverpool began by reviewing the work being done by a number of their peer Core City authorities across the country at the time. Following further investigation into the work being done at Birmingham City Council they approached Sentinel Partners. After a short evaluation and review period, Sentinel were commissioned to provide their Integrated Data Platform software, which was quickly configured to fit with all the data feeds available at that time.

Their new Data Integration Platform was commissioned at the end of August 2018 and implemented by mid-November 2018.  

As a result of their incredibly successful data integration journey, they are often cited as an exemplar authority by government departments such as the DfE and the Dept for Levelling Up, Housing and Communities (DLUHC, formally MHCLG) for their advanced uses for Data in providing more targeted and affective services.  See the related government blog at:
https://supportingfamilies.blog.gov.uk/2021/04/28/tackling-the-covid-19-impact-with-data-in-liverpool/  

Liverpool are often called upon to run or support Government online seminars and workshop sessions to help share best practice with other authorities across the country. They have also taken part in national conferences to share their experiences with the wider UK Public Sector, such as the National Data-4-Good conference presented by UK Authority:  
https://www.ukauthority.com/media/8505/liverpool-cc-sentinel-partners-how-to-master-your-data_the-liverpool-evolution.pdf  

Both the city and the council of Liverpool are recognised for their innovative, creative, and forward-thinking culture, and we are proud to be their Data Integration partner. 

SAFE Taskforce and Vulnerable Pupil Portals

The Department for Education (DfE) has allocated £3.7million of funding over a 3-year period for the SAFE (Support, Attend, Fulfil, Exceed) Taskforce to invest in and commission school focused interventions to reduce involvement in serious violence and improve attendance at schools across Liverpool. It will also help with behavior in school and within the community and improve social and emotional regulation and wellbeing. The SAFE Taskforce is a group of mainstream secondary schools working together with other multi-agency structures and local experts to support young people at risk of serious violence and re-engage them in their education. The initiative is led by schools and bringing headteachers together to support vulnerable young people and prevent involvement in gangs, county lines, and criminal activity. Attendance, reduced exclusions, and keeping pupils focused on education were key aims. The SAFE Taskforce and providing violence reduction learning in schools are some of the ways Liverpool are supporting young people in education settings across the region. 

For Liverpool we provided a Vulnerable Pupil Portal (VPP). This solution was based on the Sentinel Data Platform and designed specifically to meet the requirements on this initiative and support and empower SAFE taskforces in protecting children from becoming involved in crime, county lines, and gang exploitation.  
 
The Vulnerable Pupil Portal integrated data from multiple sources, systems and agencies to provide a complete picture of how best to ensure the appropriate level and mix of alternative provision and specialist support is delivered. Including teams from schools, police and NHS, Liverpool has portals for council, school, and police as well as Child and Adolescent Mental Health Services (CAMHS). 

Features included:  

• Dashboards specifically designed in collaboration with several local authorities displaying information and criteria needed to identify children at risk of being vulnerable 
• Pupil advanced search functionality and drill downs into single view of pupil.   
• Centralized notes enabling the sharing of information across settings and agencies. Crucial for handovers such as, change of schools and case workers.  
• Automated user alerts, including behavioral, attendance, and Police intelligence. E.g., police related incidents can alert certain users and provide contact information for further information.  
• Local Authority Vulnerable Pupil Portals included provision to produce evaluation reports and submissions to DfE.  
• Police Portals with specifically configured dashboards, functionalities, and access remits.   
• Child overviews containing all information concerning a child collated in one easy to view screen. Specific child with actions dashboard function including Safeguarding data, Attendance/Exclusions data, SEND data, Exploitation Risk, Youth Offending Team data feeds, Police data, Social Care plan, Alternative Provision mentor notes, and full workflow and chronology.  
• Additional information containing ‘soft intelligence’ such as Police cautions, behavioral and teacher notes.  
• Workflow and Chronology. Displaying a full list of past and upcoming actions and events including tracking of attendance and case worker actions as well as changes reflecting increased or decreased risk, progress or regression. 

See more detail our Vulnerable Pupil Portal solution for SAFE taskforces and Alternative Provision (AP) settings  

Research Project: Identifying future vulnerability of Domestic Abuse

A Crime Survey for England and Wales showed that an estimated 2.3 million adults aged 16 to 74 years experienced domestic abuse in the twelve months to March 2020 (1.6 million women and 757,000 men). During the coronavirus pandemic there was a 7% increase in police recorded offences flagged as domestic abuse-related between March and June 2020. In Liverpool, the number of domestic abuse cases has risen by 9% when comparing April to December 2020 to the previous year. There has been a 10% growth in the number of child protection plans which cite emotional abuse as a core referral reason since the start of the COVID-19 pandemic. 

Working with Liverpool City Council's Children's Social Services and associated Data Teams, we planned a research project based upon the huge, unified dataset we created for them in their Sentinel Data Platform. The objective of the project was to apply the latest Microsoft AI Machine Learning technologies to gain an understanding and risk measure of the degree of vulnerability for a person becoming a future Domestic Abuse (DA) victim or perpetrator.  

We had already been brought in to help Liverpool manage the city’s response to the COVID-19 pandemic by creating an index of vulnerable residents based on 45 data feeds. Information on domestic abuse arrests, interventions and callouts is regularly captured from Merseyside Police and Liverpool City Council, which is used to identify complex families as one of the six pillars of the Troubled/Supporting Families Programme. Domestic violence is a presenting factor in 25% of identified Troubled Families. Children living in a home where domestic abuse occurs can have a significant impact on a young person's mental health, physical wellbeing and behavior which can last into adulthood. 

Liverpool's Data Platform was already receiving and integrating data from 72 data sources, including all core council systems, and all strategic partner agencies they work with to support children and young families. The Platform provided a Single View of all people across the city, and this information could be rolled up into family compositions to show a Single View of each family.  
 
With Liverpool’s Corporate Intelligence Team we conducted a unique project to identify contributory factors and understand why persons become domestic abuse perpetrators based on historic and current data held within the Liverpool Families Programme data system. The key objectives were to:  

• Identify trigger reasons for an individual becoming a perpetrator of domestic abuse using scholarly literature and the judgement of professionals. These could include identifying historic household abuse (identifying if this was prevalent when a current perpetrator was a child), adverse childhood experiences (ACE’s), previous neglect, school performance, substance misuse and mental health amongst others. 
• Score the trigger reasons and overlay these over cohorts of individuals to understand if there are standout environmental or learnt behaviours that are prevalent in domestic abuse perpetrators. 
• Develop data timelines. We worked with Liverpool’s Corporate Intelligence (LCIT) Team to develop a new test system with multiple dashboard screens to identify trend data and model this using test scores. LCIT anticipated that there could be as many as 6 or 7 models created as part of the pilot to identify individuals who may have indicators or patterns in the data that suggest they could be vulnerable to becoming domestic abuse perpetrators. 
• Envisage that the modelled data and risk factors gathered during the pilot will form part of Early Help assessments to ensure that families—which are displaying typical environmental behaviours which could lead to future violence—are addressed by professional services. 
• Use data from the pilot strategically as part of the Demand Management process to identify ‘hotspot’ areas of potential future domestic abuse offenders in the city. Based upon this understanding, larger scale interventions could be implemented (e.g., in a school setting to educate young people on the risks associated with domestic abuse).  
• Learn from the pilot study to develop further risk management tools for families in the Sentinel system to better understand risk-based problems in other aspects of complex families (e.g., child exploitation). 

The data was analysed to show all current and historic issues, needs, and dependencies. The analyses from this exercise could then be used to target Family Support Services and Early Help. For the key source system, historic data was also incorporated into the Platform, providing approximately 10-12 years history of incidents, events, and outcomes.  
 
It is in this scenario where the Data Platform excels. Analyses, insights, modelling and predictions are only as good as the data. In terms of AI Machine Learning technologies, the broader, richer and deeper the information fed in, the better and more accurate the analyses. The Data Platform allowed Liverpool to pool vast amounts of complex and valuable information so that Machine Learning processes can effectively identify those at risk of becoming a future Domestic Abuse (DA) victim or perpetrator. 
 
From December 2021 to May 2022, we worked with the Liverpool teams to agree a baseline set of background issues and trigger points that current DA victims and perpetrators have in common, listed by degrees of commonality. We then used this baseline to apply Machine Learning modelling and build up confidence models of how much weighting the different baseline factors had, and where other previously un-recognised background factors also had notably high frequencies. This modelling was then applied so a subset of the wider population to show how young people with a higher potentially vulnerability could be identified.  

The objective was to show how Liverpool's unified data asset could be used to drive an early intervention strategy and support vulnerable people before they reach a point of crisis. We produced a Empowering MDM With Efficient Data Matching in the Cloud White Paper with the Liverpool teams that would be used as the basis for a business case across the council and key partners to obtain funding for a full implementation of this technology to drive proactive support services. The White Paper showed evidence of the research results, including an analysis of confidence ratings for all factors considered, and described how this approach and technology can be applied to many other scenarios such as future vulnerability of being a Looked After Child.