Design & implement a secure AI integrated system

I am seeking an experienced Microsoft 365, Power Platform and AI integration specialist to design and implement a secure Executive Operating System.

The objective is to create one reliable system for managing:

Strategic decisions and decisions required
Actions and priorities
Delegated actions and Waiting On items
Topics, projects and workstreams
Risks, opportunities and ideas
Supporting documents and evidence
Meeting, email, voice-note and transcript capture

I want one authoritative master record with clear links between Decisions, Actions, Topics and supporting evidence.

The finished system should allow me to:

Capture information through ChatGPT, Microsoft Copilot, email, meetings, mobile input and 'Looki' voice-note integration
Create and update actions using natural-language prompts
View and complete daily actions through one practical interface
Automatically update the authoritative record when actions are completed
See decisions awaiting approval, priorities, overdue actions, Waiting Ons and implementation progress
Access the system securely from desktop and mobile
Avoid manually updating multiple systems

My preferred environment is Microsoft 365, but I am open to the best justified architecture. The specialist should assess options including:

SharePoint Lists
Dataverse
Power Apps
Power Automate
Microsoft To Do
Planner
Power BI
Microsoft Graph
Copilot Studio
Secure ChatGPT API or custom integration
Looki API integration

Some existing Excel workbooks, Microsoft To Do lists, SharePoint assets, OneDrive folders and Power Automate flows already exist. These must be inventoried and preserved initially, but the final design does not have to retain the current architecture if a more reliable and maintainable solution is recommended.

I want to wake up and ask my AI agent for a morning brief. I want it to be able to tell me what's on my calendar, what are my priorities and any actions due today. I want to be able to provide a single update on actions whether that's direct to the AI agent; or completing the action on 'TO-DO' or creating a voice note on LOOKI L1. I then want the single register updated accordingly so that the master source document knows the current status and is therefore reflected across all touchpoints (listed previously). I want to be able to leave a voice transcript with ideas and decision points and then this be added to a decision register with any applicable action items branched off this. I also want to future proof it so that we can add abilities on, in the future. I want my AI agent to be able to write and read specific documents in my onedrive which will act as the 'procedures' and framework for operating and decision making etc