Durham History Centre - Collection Integration and Artificial Intelligence

Description

Summary of the work A partner is required to design and implement a cloud-based solution for ingesting and managing historic collections data from a range of existing systems, harnessing AI/ML technologies to enrich our data, and implementing appropriate unified endpoints for surfacing the data (API's) for use in websites and similar applications. Latest start date Monday 10 January 2022 Budget Range Maximum budget available £115,000, including all expenses Why the Work is Being Done The history centre will bring together multiple services and collections from across Durham into one building, creating new ways for people to engage with heritage, learn, explore, and be inspired. It will be a central gathering point for memories, stories, and knowledge about the County, creating a legacy for the future. Digital is an essential part of the history centre experience – for visitors to the new building, researchers and those planning a visit, and those who may never visit. Digital will play a fundamental role in: • Bringing together a range of heritage collections and services for better access and engagement, creating something that is bigger than the sum of its parts. • Creating an in-person experience that's intriguing, inspiring, tempting visitors to learn and engage more deeply. • Building an off-site offer that's exceptional, giving people ways to engage even if they are not physically present. • Raising the profile of Durham nationally and internationally. • Developing opportunities for cross-selling (both for revenue and engagement) between and across services. In the process of bringing together the collections, emerging technologies such as cognitive search, computer vision and artificial intelligence will be harnessed to improve data matching and metadata tagging. Problem to Be Solved You will be required to work with ourselves to process collection datasets, providing a mechanism to import datasets into our cloud environment. Once the datasets are available, they should be de-duplicated, tagged using AI and a searchable index created. AI, machine learning and cognitive search capabilities should be harnessed to improve the quality of search results and of our underlying collections by identifying similar records. Citizens should be able to search across all our collections seamlessly. This will include support for a variety of endpoints including our new website, hand-held devices, and interactive displays. It should be possible for us to share information with partner organisations using these endpoints. Alongside this the project team should be shown how to carry out these tasks so they can continue to import and process datasets beyond the life of this procurement. Who Are the Users As a member of the public, I want to be able to search across all the council's collections easily and intuitively using my chosen device so that I can discover the information that is relevant to me regardless of which collection it originates from. This will include audio-visual, geospatial and textual information. I also want to receive tailored recommendations for other items which I may be interested in, based on my search history, location or the date. As a project team application developer, I need to understand and use the appropriate technologies that allows me to provide a solution to our residents. As a collection owner, I need to be able easily to add to and update the information and artefacts available to our users through my existing systems and have confidence that this information will be processed by our new cloud solution using AI. Early Market Engagement In August/September 2021, proof-of-concept projects were carried out with both Microsoft and AWS partners, using a subset of the data within our collections, to help us de-risk the project through being able to see AI technologies in a real life scenario. The outcomes of this work have informed these requirements and given us confidence that a working solution can be obtained that meets our requirements. Work Already Done Default AI/ML algorithms were found, on first pass, to produce quite accurate tagging of object types, although some terminology was inappropriate, whilst other tags were irrelevant or repetitive, so we would be seeking ways to customise these in the main project. Text recognition and sentiment analysis of audio recordings was again found to be generally accurate. Identification of duplicate and similar images was useful and gave rise to "related results" within the sample user interface. Existing Team The supplier will be working with: • IT Project Manager. • Senior IT Officer • 4x Application Developers (approx) • Collection Leads from the five services • Digital Engagement Officer to be appointed early 2022 The procured partner and the council's Applications Development team will work together so that knowledge and skills transfer can take place, allowing the ongoing support and development of the collections beyond the current project Current Phase Alpha Skills & Experience • Experts in cloud infrastructure specifically relating to AI, cognative searching and the appropriate cloud services that are required to allow us to deliver an innovative solution • Experience in providing a solution that is innovative but also cost effective, utilising cloud solutions to help provide both, and ensuring the costing envelope for ongoing monthly SaaS remains affordable • Migration of legacy relational, flat-file and web-based content into appropriate cloud-based and cost effective storage technologies • Development of processes to enable users to import and maintain stored content. • Use of AI for processing images, text and audio content, including refinements to models which result in accurate and relevant classification. Use of AI to identify related items within datasets. • Query optimisation to ensure acceptible performance of user operations meaning search results can be provided in 3-4 seconds • Development of web services and API’s to allow collections information to move seamlessly between systems and endpoints. • Experience of delivering solutions which comply with the Public Sector Bodies Websites and Mobile Applications Regulations. Nice to Haves • Working with heritage collections • Experience of identifying gaps in the knowledge and skills of a client team and in identifying and supporting appropriate learning opportunities to fill these • Working with similar sized projects with the size and complexity of the number of collections and datasets that form part of this project Work Location Durham County Council, County Hall, Durham, DH1 5UL Working Arrangments You will work alongside (using Microsoft Teams) the council's developers and collections staff to ensure maximum understanding, acceptance and knowledge transfer throughout the project. The council's normal business hours are 8:30am to 5:00pm Monday to Friday, excluding English public holidays and Christmas closure. You may be expected to attend occasional on-site face-to-face meetings with the Durham project team, for which travel and accommodation arrangements should be made through the council's standard process unless agreed otherwise. Security Clearance Consultants will conform to our third-party access policies to allow access to any data/ systems currently held by the Council. If data is required to be consumed in a test system held off-site by the consultant, it should be removed and destroyed within a 30 day period upon contract end. Additional T&Cs We will require evidence of Professional Indemnity insurance cover. No. of Suppliers to Evaluate 5 Proposal Criteria • Demonstrate how your knowledge and experience will help Durham to maximise the value of our digital collections through use of AI/ML/cognitive search and related technologies, including two relevant case studies • Demonstrate how your solution will enable our residents and visitors both to find the information they need and to explore our collections without having a specific goal in mind. • Describe the range of people you will bring to the project and how their skills and experience are relevant to its aims. • Demonstrate how your approach to the project will result in the transfer of knowledge to, and development of skills of, our in-house team so the product is supportable and sustainable. • Explain how your approach will ensure all stakeholders' expectations are understood and reflected in a unified and agreed design and how you’ll ensure this is delivered on time and budget. • Deliver value for money. Cultural Fit Criteria • Working with multiple stakeholders with potentially competing views • Challenging existing ways of working/the status quo • Working closely and collaboratively, sharing skills, coaching and supporting to ensure a high-quality and sustainable outcome • Monitoring emerging developments in the underlying technologies throughout the project and ensuring these are utilised as appropriate • Putting the customer experience at the heart of the solution • Working with the project team to ensure they are fully equipped to carry out the project using technologies implemented. Payment Approach Capped time and materials Assessment Method • Case study • Work history • Reference • Presentation Evaluation Weighting Technical competence 60% Cultural fit 15% Price 25% Questions from Suppliers 1. Will you reject applications of they do not have ‘Working with heritage collections’ experience ? If yes then it will help us save time and resources. As part of the shortlisting process the Council can exclude Suppliers who have the least nice to have skills and experience in order to arrive at the required number of Suppliers to evaluate. 2. Would there be a need to provide a written proposal at this stage or it is required only after the suppliers have been shortlisted? A written proposal will only be required by those suppliers who are shortlisted. 3. Would there be multiple phases within which the solution will be developed? The current phase is mentioned as ‘Alpha’. Would there be a ‘beta’ phase, for example? The alpha stage, where we currently are, will see us building solutions and testing them, before moving into a beta/ pre go live. It is expected that the alpha stage continues through most of 2022. It is likely that what is build as part of alpha will become the beta solution. We will need to have a working solution for end point consumption early 2023. 4. What does the budget cover? For example: does the budget cover the professional services costs, cloud consumption costs, licensing costs etc? In addition, what is the timeline horizon for the costs e.g. delivery timeline? The budget should cover all costs incurred as part of the implementation of the system. This should include cloud consumption/ licensing etc. Costs beyond implementation are not covered.For the timeline horizon we would expect to start incurring costs as soon as a tenancy is created, and services activated. This will be relatively quick in the project as we have lots of collection items to pass through the cloud service. This could, however, be in the form of a development environment but we would need to work with the successful organisation to determine what’s best for the project delivery.