Stotles logo
Awarded

Open Leo Data Service -DfE

Published

Supplier(s)

Advanced Skills Initiative Ltd

Value

750,000 GBP

Description

Summary of the work Department for Education requires a supplier team to lead on development and delivery of a technological solution to enable open access to Leo dataset whilst preserving individual anonymity. Closest delivery approach would be Beta with focus on MVP only. Expected Contract Length 9 months Latest start date Friday 1 July 2022 Budget Range The services are to be delivered over a contract period of 9 months. The indicative value for the full 9-month contract is £750,000. Suppliers shall be asked to provide quotation for each stage of the required work, with an exit option at each stage, should we choose not to proceed. Each SOW will outline the work required, associated budget and payment approach, CTM, T&M etc. Suppliers will need to provide clear costs to enable tracking. DfE does not commit to any minimum or maximum spend at this point. A pricing template will be provided to shortlisted suppliers at proposal stage Why the Work is Being Done The Department for Education's newly formed Unit for Future Skills (UFS) is tasked to lead a cross-Government programme to transform how jobs and skills data is used to inform learners, training providers, and policymakers and support a responsive skills ecosystem and enable a wider change in the use of data in DfE policymaking and delivery. To facilitate the delivery of these aims, the Unit’s ambition is to develop data infrastructure and new online product(s)/service where the jobs and skills data can be brought together, easily visualised and made securely available to individuals and institutions operating in jobs and skills markets with ability to share underlying data where possible with our users or via third-party service providers. Problem to Be Solved The longitudinal educational outcomes (LEO) dataset forms the backbone on which most of Units analysis and insights rest, it has shown itself to be a significant data resource to provide insight in the supply of skills across the country and their value in the labour market. The current channels for sharing the LEO data are limited and restrictive. Widening the access to the LEO dataset is part of the DfE broader strategy of making data more available, to increase the potential for insights to be developed and acted upon. To facilitate the delivery of these aims, the Unit’s ambition is to set-up Open Leo data service to bridge this gap, making granular data available in accessible formats for all our users, with flexibility to query, extract and visualise data to meet needs for all our users, whilst preserving individual anonymity and compliance with departmental privacy requirements for publishing and sharing of data in the public domain. Who Are the Users The service will be used by broad range of users, which will include but not be limited to learners and their parents, adult learners/citizens, career advice services, employers, and employer representative bodies (ERBs), MCAs/local bodies, providers, research and academic community, internal DfE skills policy colleagues and private organisations interested in pathways between skills, qualifications and employment outcomes Early Market Engagement N/A Work Already Done The department has undertaken some work previously on sharing Leo dataset. We would like the service providers to work together with department to build upon and validate user insights in support of delivering MVP. Existing Team "There is no existing team for this work. The DfE Teams will include Head of Data and Digital Services/ Service Owner, Product Owner and analytical team members who will agree SoW and sign-off on the deliverables". Current Phase Beta Skills & Experience • Expertise in successfully developing, testing, and deploying automated disclosure or privacy preserving algorithms (e.g., differential privacy) on complex datasets • Demonstrable experience in designing and delivering end to end scalable cloud-based data science, machine learning/ artificial intelligence solutions including tooling within government or private sector. • Experience in deploying easy-to-use dashboards, apps, and data visualisation tools • Capability to implement machine learning data models utilising links to open data sets and model libraries. • Good understanding of data privacy regulations for publishing data in public domain • Capability to build and deploy models and tooling using departmental infrastructure. Nice to Haves Experience and understanding of educational datasets and digital services Work Location Mainly Remote working. Supplier to attend in person for work progress update, user group, and contract management meetings in DfE’s London and Manchester Piccadilly office as and when required. Working Arrangments Department for Education requires a supplier team to lead on development and delivery of a technological solution to enable open access to Leo dataset whilst preserving individual anonymity. The specialist supplier will have expertise in development and testing of automated disclosure control / privacy preserving algorithm (e.g. differential privacy) on complex datasets. We expect the services will be outcome based with pre-agreed deliverables for each stage as per agreed statement of work. The work will be time and cost capped. Security Clearance The successful supplier must be able to demonstrate that all proposed team members have been subject to Baseline Personnel Security Standards checks. Suppliers will also be required to complete a supplier security assurance form to ensure they meet the required standard for DfE e.g. Cyber Essentials Additional T&Cs Standard Framework and Call Off Terms and Conditions. Expenses must be pre-agreed and comply with CCS Travel and Subsistence Policy. Any expenses shall be submitted in line with DfE standard T&S policy. Primary work location stated in SoW will not attract expenses. Contract and Vendor Management will form a key part of governance and suppliers will be expected to complete cyber security questionnaire. Suppliers must provide sufficient guarantees to meet the requirements of GDPR in line with Procurement Policy Note 03/17 Changes to Data Protection Legislation & General Data Protection Regulation No. of Suppliers to Evaluate 4 Proposal Criteria • Expertise in successfully developing, testing, and deploying automated disclosure or privacy preserving algorithms (e.g., differential privacy) on complex datasets • Demonstrable experience in designing and delivering end to end scalable cloud-based data science, machine learning/ artificial intelligence solutions including tooling within government and/or private sector. • Experience in deploying easy-to-use dashboards, apps, and data visualisations • Capability to implement machine learning data models utilising link to open data sets and model libraries, so as to avoid use of proprietary models. • Good understanding of data privacy regulations for publishing government data in public domain. • Capability to build and deploy models and tooling using departmental infrastructure. Cultural Fit Criteria • Experience of transferring knowledge to permanent staff within a client organisation • Experience of working within multi-vendor teams • Describe how your organisation encourages diverse representation of under-represented groups in the workforce, e.g. Women, Black, Asian and Minority Ethnic, Disabled, LGBTQ+ and how you monitor and measure this (10%) Payment Approach Capped time and materials Assessment Method • Case study • Work history • Reference • Presentation Evaluation Weighting Technical competence 50% Cultural fit 20% Price 30% Questions from Suppliers 1. Please confirm who will be on your evaluation panel The panel will include Statistician, Head of Analysis, Head of Data & Digital Services /Service Owner, Unit for Future Skills and Data DirectorateAs part of the evaluation, the panel may seek advice from subject matter experts and Heads of Profession in data architecture, engineering, operations, data science and statistics as needed. 2. Does this requirement sit under DFE to aid all ALB’s linked to the department or will this be owned and managed by one of the entities within the DFE department? The relationship with ALBs will be primarily owned and managed by DFE. The selected service provider will be expected to engage and manage any specific work strands including but not limited to user testing with ALBs as identified as part of the statement of work. 3. Can you confirm which supplier built the leo Data set? LEO is a de-identified, person-level administrative dataset and is jointly owned and developed by DFE, DWP and HMRC. 4. What tech stack was used for the Beta service for LEO The beta (MVP) has not been developed yet. Currently, DfE publishes aggregate LEO data is through Explore Education Statistics platform, which utilises SQL databases hosted in Azure and C#/.net backend. The full repo is available https://github.com/dfe-analytical-services/explore-education-statistics. 5. Is the work outside or inside IR35 This work scope is outside IR35 6. Has DfE have a preference for hosting of the service e.g. MS Azure or GOV.UK PAAS ? We would like the service provider to work together with department to evaluate the existing hosting options, with an aim to select the most efficient and cost-effective route to host this service. See https://dfe-technical-guidance-135.london.cloudapps.digital/infrastructure/ 7. Do you currently have a view of which subset of the fields in LEO you might want to publish? If so, can you give guidance on this prior to submission? The data fields will be dependent on use cases selected for user testing and will be agreed with the selected service provider. 8. Do you require the supplier to select and design disclosure control algorithms in addition to implementing and testing them? Yes 9. Has there been any prior work performed on disclosure control algorithm selection and design? No 10. Has an alpha system been built? If so, what technology stack is used? No. Currently, we share aggregate LEO data is through Explore Education Statistics platform, which utilises SQL databases hosted in Azure and C#/.net backend. The full repo is available https://github.com/dfe-analytical-services/explore-education-statistics. 11. Has any research been conducted into user needs in relation to opening up the LEO dataset? The department has undertaken some work previously on sharing LEO dataset. We currently publish aggregates from LEO dataset and have identified an increasing demand to make more granular data accessible via APIs and dashboards. We would like the service provider to work together with department to build upon and validate user insights in support of delivering MVP. 12. Have you done any research into comparable work done in other countries? No 13. Could you add some detail or examples of departmental infrastructure? For development of departmental products and services, we utilise infrastructure as code on Azure with CI\CD pipelines, for databases Microsoft SQL Server or Cosmos DB for preference with Azure Function RESTful APIs to split service layers and Redis caching if required (with authentication layer if needed). Software is cloud first, C# .NET (Core).Data Engineering Tools - Informatica DEI, Informatica Data Quality, Azure Databricks, Azure Data Factory, Python Storage - ADLS Gen 2, Azure Blob Storage, Cosmos DB, Microsoft SQL server. 14. How is the LEO dataset currently stored and accessed? What is the data model, the database, the APIs? Leo dataset is currently stored and accessed using Microsoft SQL server. Currently, there APIs are not being used for LEO dataset.Selected service provider personnel will be provided access to LEO dataset via DfE laptops to develop, test and prove application of the privacy preserving algorithms. 15. Are Linked Data and RDF Knowledge Graphs acceptable as data infrastructure for the Open Leo service? Currently RDF knowledge graphs are not widely used. See examples of acceptable infrastructure in Q13. We would like the service provider to work together with department to utilise our existing data infrastructure as we would need to consider the ongoing maintenance, support and lifecycle costs including availability and cost of support. 16. What kind of work has already been done? The department has undertaken some work previously on sharing LEO dataset. DfE publishes aggregates from Leo dataset on the Explore Education Statistics website and more granular de-personalised individual data is available to accredited researchers via ONS SRS upon application. This project is departmental first on application and testing of automated disclosure control on LEO data set. 17. How will the ML/AI solutions be used? Service providers can recommend privacy preserving/ automated disclosure control solutions, which may use ML/AI. 18. Does the budget include infrastructure costs such as a commercial database license (which is usually annual)? The budget includes the infrastructure costs for the period of development and testing of MVP over the duration of 9 months. 19. Will Open Leo be an internal government service or a public one? Open Leo is envisaged to be a public service. 20. What kind of infrastructure will the solution have to be deployed to? For development of departmental products and services, we utilise infrastructure as code on Azure with CI\CD pipelines, for databases Microsoft SQL Server or Cosmos DB for preference with azure function RESTful APIs to split service layers and Redis caching if required (with authentication layer if needed). Software is cloud first, C# .NET (Core).Data Engineering Tools - Informatica DEI, Informatica Data Quality, Azure Databricks, Azure Data Factory, Python Storage - ADLS Gen 2, Azure Blob Storage, Cosmos DB, Microsoft SQL server. 21. What kind of SLA is planned for the solution? SLA will be agreed with the selected service provider. 22. How is the LEO dataset currently stored and accessed? What is the data model, the database, the APIs? Leo dataset is currently stored and accessed using Microsoft SQL server. Currently, there APIs are not being used for LEO dataset.Selected service provider personnel will be provided access to LEO dataset via DfE laptops to develop, test and prove application of the privacy preserving algorithms. 23. Are Linked Data and RDF Knowledge Graphs acceptable as data infrastructure for the Open Leo service? Currently RDF knowledge graphs are not widely used. See examples of acceptable infrastructure in Q13. We would like the service provider to work together with department to utilise our existing data infrastructure as we would need to consider the ongoing maintenance, support and lifecycle costs including availability and cost of support. 24. What kind of work has already been done? The department has undertaken some work previously on sharing LEO dataset. DfE publishes aggregates from Leo dataset on the Explore Education Statistics website and more granular de-personalised individual data is available to accredited researchers via ONS SRS upon application. This project is departmental first on application and testing of automated disclosure control on LEO data set. 25. How will the ML/AI solutions be used? Service providers can recommend privacy preserving/ automated disclosure control solutions, which may use ML/AI.

Timeline

Publish date

2 years ago

Award date

2 years ago

Buyer information

Explore contracts and tenders relating to Department for Education

Go to buyer profile
To save this opportunity, sign up to Stotles for free.
Save in app
  • Looking glass on top of a file iconTender tracking

    Access a feed of government opportunities tailored to you, in one view. Receive email alerts and integrate with your CRM to stay up-to-date.

  • ID card iconProactive prospecting

    Get ahead of competitors by reaching out to key decision-makers within buying organisations directly.

  • Open folder icon360° account briefings

    Create in-depth briefings on buyer organisations based on their historical & upcoming procurement activity.

  • Teamwork iconCollaboration tools

    Streamline sales workflows with team collaboration and communication features, and integrate with your favourite sales tools.

Stop chasing tenders, start getting ahead.

Create your free feed

Explore other contracts published by Department for Education

Explore more open tenders, recent contract awards and upcoming contract expiries published by Department for Education.

Explore more suppliers to Department for Education

Sign up