Artificial intelligence services
Microsoft 365 Copilot, private assistants and AI automation.
AUL deploys Microsoft 365 Copilot, private and local assistants, custom agents and workflow automation. We integrate AI with your documents, applications and knowledge bases, with hosting, permissions and human review defined around the chosen use case.
AI use cases we implement
A concrete use case makes it easier to assess whether AI is worth introducing.
- You are considering Copilot and want to assess your Microsoft 365 environment first.
- Staff spend time searching internal documents or answering repeated questions.
- A repetitive process could benefit from automation with human review.
Five distinct AI services
From a Microsoft 365 rollout to fully local AI, we implement the approach that fits the task, data and operating environment.
Microsoft Copilot deployment
We roll out and configure Microsoft 365 Copilot across your tenant around selected business data and high-value use cases. Copilot Studio can extend the solution with tailored assistants and integrations.
- Configure Microsoft 365 Copilot for the tenant and the selected users and use cases.
- Connect relevant data and configure Copilot Studio extensions for the agreed solution.
Platforms and tools: Microsoft 365 Copilot, Copilot Studio
Private and local AI assistants
We build private assistants grounded in your internal documents and knowledge bases, including fully local installations inside your own environment. Model hosting, document access and any external connections are configured for the chosen deployment.
- Connect approved internal documents and knowledge bases to the assistant.
- Deploy the assistant locally or in the selected private environment and configure access and external connections.
Custom AI agents for staff and customers
We design and deploy agents that answer questions or carry out defined tasks for employees or customers. The solution combines access to relevant systems with task boundaries and human oversight.
- Connect agents to the business applications and data sources required by the selected tasks.
- Configure task boundaries and human review points for staff or customer interactions.
AI workflow automation with human approvals
We automate repetitive, multi-step business processes by combining AI with the systems you already use. Human approval checkpoints are built into the workflow where actions need review.
- Integrate AI steps with the existing applications and data used by the business process.
- Configure human approval checkpoints before the selected actions proceed.
On-premises and sovereign EU cloud AI infrastructure
We deploy AI on your own infrastructure or in a sovereign EU cloud environment. Hosting and connectivity are designed around your data-control requirements and the operating needs of the solution.
- Set up the selected on-premises AI infrastructure or sovereign EU cloud deployment.
- Connect deployed models and assistants to the agreed data sources, applications and knowledge bases.
Three ways to work with AUL
Get advice on the approach, end-to-end integration or ongoing care for the AI solutions your business uses.
Advisory and readiness support
Select the right cloud, local or hybrid AI approach for your goals and data. We provide governance advice and readiness support for GDPR, NIS2 and the EU AI Act requirements relevant to the solution.
Integration and deployment
Deploy and configure Copilot, private assistants, agents or automation, and connect them to existing data, applications and knowledge bases. We define hosting, access, data flows and human approval points for the implementation.
Monitoring, updates and optimisation
Monitor, update and optimise the AI solution as your needs change. Ongoing support addresses performance, output quality and operating costs across the agreed systems and integrations.
Decisions to discuss before work starts
- Which task, users and data sources are in scope?
- Where will the model, documents, search indexes and logs be hosted?
- Which integrations, permissions and human approvals are required?
- How will usefulness, output quality, cost and ongoing ownership be evaluated?
Questions about these services
How do we choose between Copilot, a private assistant and an AI agent?
Start with the task, the data it needs and the applications it must use. Copilot supports Microsoft 365 workflows; a private assistant can connect to internal documents and knowledge; a custom agent can perform a defined task with agreed permissions and human review.
Can an AI assistant run on our own infrastructure?
Yes. AUL offers local and private AI deployment, including on-premises, cloud and hybrid options. We define the compute, storage, access and integration requirements around the chosen models and use case.
Related services
Which task would you like AI to help with?
Tell us about the task, the tools you use and the data involved. We can discuss which approach is worth assessing.