Practical AI Implementation
Case Study: Briefable’s Automated Morning Briefing System
How an automated morning briefing system built on Model Context Protocol synthesizes daily schedules, client deal histories, and inbox threads without human copy-pasting or manual data preparation.
Google Calendar MCP Server
Queries daily attendee lists, event metadata, and meeting notes directly
Gmail Context MCP Server
Retrieves recent email correspondence threads related to specific participants
Supabase CRM Database Server
Fetches active commercial deal stage, contract status, and historic notes
Synthesis & Structuring Engine
Generates prioritised daily execution prep notes delivered at 06:45 AM
Executives and consultants enter morning meetings fully briefed on client history without spending time reviewing separate software applications.
Zero manual data copying required between systems
Adheres strictly to the open Model Context Protocol standard
Provides complete local data logging and permission isolation
Operational Context
The daily meeting preparation bottleneck
For executive leaders, senior consultants, and project directors across New Zealand, the first hour of every business day is often consumed by administrative gathering. Managers must open Google Calendar, review attendee names, cross-reference contacts against active CRM pipeline deals, search Gmail for recent conversation threads, and recall outstanding contractual obligations.
In mid-sized professional services firms, this administrative ritual costs forty-five minutes per professional every morning. When calculated across an operational team of twelve people, the firm loses over twenty-five hours of billable senior capacity every week. Previous attempts to solve this problem involved hiring junior executive assistants or building complex custom software integrations that broke whenever vendors modified software endpoints.
Briefable, an operational SaaS system within the Changeable ecosystem, addressed this exact friction. Rather than building fragile API connectors or relying on staff to copy-paste context into web chats, Briefable implemented an automated morning briefing system built entirely on Model Context Protocol architecture. This deployment aligns directly with the practical AI education published on Zero to AI guidance channels.
Fragmented client intelligence across software
Meeting records live in calendars, commercial status lives in CRMs, and project risks live in email inbox threads. Operating across disconnected systems results in missed context during client conversations.
Low return on manual preparation routines
Senior leaders spend valuable early-morning focus performing manual information lookup rather than high-value strategic decision-making and business development.
Data privacy and credential exposure risks
Copying client correspondence into unvetted consumer AI portals exposes commercial terms and breaches organizational data governance standards.
How the Briefable automated morning briefing system operates
A breakdown of the native protocol connections enabling real-time synthesis across calendar, CRM, and communication channels.
Automated Schedule Query
At 06:30 AM, the system triggers a query via the Google Calendar MCP server, retrieving all scheduled meetings, attendee email addresses, and event descriptions for the day ahead.
Real-Time CRM Matching
The AI model queries the Supabase CRM database via an internal MCP server, matching attendee domains against active proposals, contract histories, and account manager notes.
Contextual Email Retrieval
Through the Gmail MCP server, the model scans recent email exchanges with meeting attendees, isolating unresolved questions, outstanding action items, and recent attachments.
Intelligent Synthesis
The core intelligence engine synthesizes the retrieved data, highlighting critical agenda items, potential deal risks, and recommended discussion points for each engagement.
Formatted Delivery
At 06:45 AM, a structured daily briefing is delivered directly to the executive’s private reading interface, organized chronologically with direct links to underlying systems.
Strict Permission Boundaries
The protocol operates strictly within the user’s explicit credential authorization, ensuring confidential board or financial matters remain restricted.
Technical Architecture & Data Flow
Four structural layers ensuring secure, high-velocity synthesis without centralizing credentials or leaking sensitive data.
Client Request Layer
The execution pipeline initiates via a scheduled cron job executing within a secure Supabase Edge Function environment.
- Triggers at designated time zones (NZST/NZDT)
- Initializes secure session token parameters
- Establishes temporary memory buffer
- Enforces local tenant isolation
MCP Server Mesh
Decentralized MCP servers expose specific system tool interfaces to the central AI reasoning engine.
- Calendar Server: exposes
get_daily_events - CRM Server: exposes
get_deal_context - Gmail Server: exposes
get_recent_threads - Strict JSON-RPC communication protocol
Reasoning & Filtering Engine
Claude 3.5 Sonnet processes retrieved context via structured system prompts, applying strict relevance filters.
- Ignores routine newsletter/marketing threads
- Prioritizes open commercial proposals
- Surfaces critical action deadlines
- Maintains neutral, factual synthesis tone
Human-in-the-Loop Interface
The finalized output presents structured prep cards with clear source citations and manual modification options.
- Presents chronological meeting cards
- Includes direct deep-links to original emails
- Allows one-click addition of personal notes
- Logs delivery completion metrics
Key Outcomes
Quantifiable impact on executive operations
Implementing the Briefable automated morning briefing system resulted in immediate operational improvements and measurable time savings across leadership roles.
Strategic Lessons
Why protocol integration beats standalone SaaS tools
The success of the Briefable build demonstrates why New Zealand organisations should focus on protocol infrastructure rather than purchasing additional isolated software applications. Conducting rigorous operational workflow mapping ensures technology investments directly eliminate administrative overhead.
By leveraging open standards, organisations retain full ownership of their underlying system connections and data flows. When new AI models are released, the protocol layer remains unchanged, protecting the firm from costly redevelopment cycles or vendor lock-in.
Context Beats Features
The value of executive AI lies in giving the model access to live operational context, not in complex prompt engineering or slick user interfaces.
Zero Clipboard Drag
Automating data retrieval eliminates human error, copy-paste fatigue, and the privacy risks associated with manual data handling.
Future-Proofed Scalability
An open protocol connection allows organizations to add new data sources, such as job costing systems, without rewriting core integrations.
Immediate Operating Relief
Delivering synthesized insights directly into existing daily routines ensures high adoption across senior leadership teams.
Questions
Briefing System Implementation FAQ
Technical and operational questions regarding the deployment of protocol-driven briefing systems.
How long did it take to build the Briefable MCP briefing system?
The core MCP server configuration and synthesis pipeline were established in less than two business days due to the standardized nature of Model Context Protocol interfaces.
Is executive email content safe when processed by the system?
Yes. Data access occurs strictly through authorized local session keys. Content is processed transiently for synthesis and is never used to train public language models.
Can this briefing system connect to Microsoft Outlook and Teams instead of Google?
Yes. Standard MCP servers exist for Microsoft 365, allowing identical briefing workflows across Outlook calendars, emails, and Teams messages.
What happens if an API credential expires or loses authorization?
The system handles credential authentication errors gracefully, notifying the user via a secure alert while maintaining existing data logs.
Can the briefing system trigger actions, such as sending emails?
While the protocol supports write capabilities, Briefable restricts the briefing system to read-only synthesis to maintain absolute Human-in-the-Loop control.
How can our organisation build a similar briefing workflow?
Changeable provides structured consulting engagements to map your internal data sources, deploy compliant MCP servers, and configure automated executive workflows.
Want to eliminate manual meeting prep in your business?
Book a Decision Clarity Session with Steve Wilson. We will evaluate your executive workflows, identify data integration opportunities, and design a custom protocol architecture for your team.