Microsoft365, Microsoft Viva, Azure OpenAI and Bot Framework consultancy

In early 2022 Microsoft approached me to find out if I would return to the University of Surrey to support a proof of concept around Microsoft Viva.

Initially, my role was to support the Gold Partner to implement Microsoft Viva in our production tenancy and role the service out across 100 members of staff.

Once the implementation period was completed and the gold partner had handed over the service. I was responsible for running the service during the proof of concept.

I was responsible for

  • Working with the finance team to procure the licenses and ensure they were assigned to the correct tenancy
  • Ensuring the gold partners pre-requisites were completed
  • Acting as the go between for the gold partner and different teams within IT services
  • Ensuring the gold partner employees had sufficient remote access limited by Azure PIM
  • Configuring Viva Dashboards
  • Building ACE cards using SPFx and the basic ACE card designer
  • Supporting the design of the Viva Dashboards
  • Running workshops to inform users about Viva Insights, Goals, Engage, Learn and how it could be used to drive employee engagement
  • Writing a summary report to highlight how SharePoint Online, Microsoft Teams and Viva could be used to deliver modern intranet services
  • Working with a separate project stream to show how SharePoint Online, Viva and ACE cards could deliver a seamless student journey
  • Writing a project closure, lessons learned and recommendations document
  • Working with Microsoft to update Microsoft Viva documentation, supporting the development of ACE cards with Microsoft PnP community

The successful Microsoft Viva proof of concept provided the foundations for implementing a replacement solution for the MySurrey mobile application. The implementation formed a key part of the Seamless Student Journey programme with a Microsoft Viva Connections experience (dashboard) for students being delivered for the start of the 2024/25 academic year.

“My Dashboard” being advertised on campus. Picture courtesy of James Pickett (University of Surrey)

The dashboard provides news, ACE cards and useful resources. The ACE cards consist of a number of custom ACE cards, which offer students information from key line of business systems including

  • E-mail and calendar events from Exchange with integration using Microsoft Graph
  • Knowledge base search based on integration with Microsoft Dynamics365
  • A summary of open support tickets with integration to Microsoft Dynamics
  • An overview of a students engagement with integration to Surrey’s student engagement platform via API’s

The dashboard also includes static Viva ACE cards built using ACE card designer to provide signposting to

  • Timetabling
  • Surrey Learn (D2L)
  • Library Search
  • Student Self Service
  • Letter generation
  • Forms
  • OneSurrey Support (Microsoft Dynamics)

OOB ACE cards have been used for

  • University Events
  • OneDrive integration
  • Planner integration

As part of the team delivering the SSJ Student Dashboard I have been responsible for

  • Implementing production, development Microsoft Viva Connection experiences dashboards
  • Working with the integrations team to establish which API gateway calls can be used
  • Working to create new integrations with partners for the engagement platform
  • Creating custom ACE cards using SPFx
  • Working with the design team and Microsoft Partners to establish the user journey, branding and design elements
  • Knowledge transfer, service transition, go live planning and documentation

Accommodation Assistant (Azure OpenAI, ChatGPT)

In addition to the Microsoft Viva development I was asked to produce an AI assistant based around my previous work at the University of Surrey and the work produced by Microsoft for the BETT 2023 exhibition. The AI assistant was used to answer accommodation based queries from students and prospective future queries.

To support this work I have

  • Implemented an Azure technology stack to support the import of content from Microsoft Dynamics and Drupal websites where content is stored in Azure Blob Storage
  • Configured Azure Cognitive Search to index Azure Blob Storage
  • Configured Azure OpenAI Studio Chat to use Azure Blob Storage, configure ChatGPT and ADA02 models, defined chat settings such as temperature and maximum tokens, configure semantic and vector search to improve search results.
  • Configured Azure OpenAI studio to use the new “bring your own data” model to limit what Chat GPT can use as a source of data
  • Created a VS Code, C#, bot framework solution and customised the code to include new ACE cards for conversation flow and enable handoff to the Microsoft Dynamics ticketing portal
  • Deployed the code to a Azure Web App using the Azure CLI
  • Deployed and configured the Azure Bot Service so that the bot is available via Web Chat and Microsoft Teams
  • Defined the JavaScript to enable the accommodation assistant to be integrated with our accommodation website
  • Completed a Data Protection Impact Assessment review
  • Supported the creation of an Ethical AI Impact Assessment and governance process
  • Supported the Information Architects in reviewing the design and implementation of the solution
  • Worked with senior stakeholders and testers to provide an understanding of how Generative AI works and explaining why the answers were given based on the questions asked
  • Configured bot services analytics
  • Worked with Microsoft to identify bugs and issues with Azure OpenAI documentation and .net modules
  • Worked with Microsoft to produce documentation and graphics for Microsoft Ignite 2023 and the European SharePoint Conference 2023
  • Worked on slide decks for stakeholders focusing on “Explainable AI”
  • Building a SharePoint communications site to introduce “Artificial Intelligence”
  • Working and managing a team of student developers who helped build and design the accommodation assistant
  • Work closely with the Surrey Institute for People-Centred Artificial Intelligence (PAI) who peer reviewed our work