top of page

Committee Process Automation: From Emails to Controlled Decisions

  • ukrsedo
  • 12 minutes ago
  • 3 min read

Most committee processes still rely on email.


Submissions arrive in different formats, agendas are rebuilt manually, attendance is guessed, conflicts of interest are tracked in people’s heads, and voting outcomes are reconstructed after the fact. The decisions may be valid — but the process is fragile, opaque, and painful to audit.


This solution fixes that.



What this automation solution is about?


This is a SharePoint + Power Automate–based committee administration solution that structures how committee submissions are:

  • registered

  • scheduled

  • attended

  • voted on

  • and formally closed


No AI aids. Full audit trail.

The system does not “decide”. It enforces discipline around how decisions are taken.


What the solution covers


The solution governs four essential parts of any committee lifecycle:

  • Entry registration and tracking: Submissions are formally registered, status-controlled, and traceable end-to-end.

  • Agenda formation: Each meeting has a single, dated, approved agenda — no moving targets.

  • Attendance and conflict-of-interest control: Voting eligibility is determined for each meeting and for each agenda item.

  • Voting and decision recording: Quorum rules are enforced, votes are counted correctly, and outcomes are recorded for each submission.


Voting can happen before, during, or after the meeting — including controlled closure when time runs out, and the agenda isn’t empty.


Flow-level design


Flow 1 — Entry Registration & Status Control

Purpose: Control the lifecycle of committee submissions.

Trigger: SharePoint item created or modified.

What it does

  • Registers new submissions

  • Detects and controls status changes (e.g. New → Scheduled)

  • Sends acknowledgements and scheduling notifications

Output: A single, authoritative list of committee entries with a clear, auditable status history.


Flow 2 — Meeting Agenda Formation

Purpose: Define what is officially on the table for a specific meeting.

Trigger: Manual run by Secretariat or Manager.

What it does

  • Identifies the next scheduled committee date

  • Selects all submissions planned for that date

  • Orders items by priority

  • Generates a structured agenda

  • Distributes it for managerial approval

  • Archives the approved version

Output: One dated agenda representing the official scope of the meeting.


Flow 3 — Attendance & Conflict-of-Interest Recording

Purpose: Define who is allowed to vote.

Trigger: Manual run with a meeting date.

What it does

  • Identifies voting members from the committee roster

  • Sends attendance confirmations

  • Records Present / Absent responses

  • Creates or updates attendance records

  • Records conflict-of-interest abstentions per agenda item

Output: A formal list of present and eligible voters for the meeting.


Flow 4 — Voting & Decision Processing

Purpose: Execute voting and record outcomes correctly.

Trigger: Manual run with:

  • meeting date

  • voting phase (pre / in / post-meeting)

  • run mode (vote or out-of-time close)

What it does

  • Selects agenda items still requiring a vote

  • Determines eligible voters per item

  • Sends approve/reject requests

  • Counts votes, abstentions, and non-responses

  • Applies quorum rules (count or percentage)

  • Handles ties via Chair vote where required

  • Writes final decisions back to each submission

  • Defers unvoted items when a meeting is closed

Output: A final decision status per agenda item, with full supporting vote artefacts.


How the flows work together

Each flow does one job — and only one:

  • Flow 1 controls what exists and when it moves

  • Flow 2 defines what is on the table

  • Flow 3 defines who can vote

  • Flow 4 defines what decision is taken

This separation is intentional. It makes the system auditable, testable, and governable.


What can be improved next?

The core is intentionally conservative. Enhancements can be layered without breaking governance:

  • AI-assisted submission critique: Flow 1 can be extended to provide structured feedback against corporate or departmental standards.

  • Better user experience: Adaptive Cards can replace standard emails for cleaner interaction.

  • Automated protocol amendments: A separate flow already exists to convert Teams meeting transcripts (VTT files) into protocol updates and action lists.


Final thought

You do not need AI to optimise processes.


Most organisational inefficiency comes from undefined states, unclear ownership, and uncontrolled transitions — not from a lack of intelligence. Those problems are solved with basic RPA and workflow design, not machine learning.

AI can enhance a stable process. It cannot fix a broken one.

Start with RPA. Add AI only after the process is under control.

bottom of page