YakYak

Fractional Chief AI Officer Proposal

For Concepcion Global P.A. - a focused plan to stabilize intake, capture lead calls, expose website data to AI search, and map firm knowledge by role.

Confidential proposal · June 2026 · $5K/month

Recommended implementation order

Opportunity Recap

These are the four concrete opportunities from the call, ranked left to right by the order we recommend implementing them.

01

Intake Command Center

Turn info@ into the reliable front door: detect new leads, reply with the $500 consult payment link, and route booking.

First build: info@ reception + paid consult reply

02

Quo Lead Call Capture

Switch the Mexico lead-caller workflow from RingCentral evaluation to a Quo pilot with local numbers and transcription.

First build: Quo pilot + transcript routing

03

Website MCP for AI SEO

Expose services, pricing, booking rules, and consult availability as structured data for Google AI discovery.

First build: pricing + booking schema

04

SharePoint Data Map

Map semantic, procedural, and episodic data handling across Admin, Attorneys, and Assistants before deeper automation.

First build: role-aware data model

Implementation 01

Intake Command Center

First scope

Info@ becomes the reliable front door.

Keep the first version tight: receive the lead, send the standardized response, and move the lead toward a paid consultation.

What it does

Monitors info@ for new leads and sends a clean reply with the $500 consultation payment link and booking link.

Why it matters

No lead should depend on manual inbox triage before the first paid consult path is offered.

First build

Lead-detection rule, approved response template, payment/calendar link, and a simple intake queue for follow-up.

Implementation 02

Quo Lead Call Capture

Phone system

Move Mexico lead calls into a recordable, transcribable workflow.

Quo becomes the pilot path for local numbers and call intelligence, compared against the current RingCentral setup.

What it does

Provides a local-number workflow for the Mexico cold-caller and captures call recordings or transcripts for lead review.

Why it matters

Lead quality should not disappear into voicemail or memory. Calls become intake evidence.

First build

Pilot Quo, confirm call-recording requirements, route transcripts into the intake queue, then decide whether to retire RingCentral for this use case.

Implementation 03

Website MCP for Google AI SEO

Discovery layer

Make pricing, booking, and consult paths machine-readable.

The website should not just rank. It should expose structured information that AI discovery systems can understand and route into booking.

What it does

Creates a structured website interface for services, pricing, eligibility, FAQ, and the paid-consult booking path.

Why it matters

Google AI and other model-driven discovery tools need explicit data, not just marketing copy.

First build

Define service/pricing schema, booking endpoint, consultation rules, structured data, and an MCP-style interface for model access.

Implementation 04

Role-Aware SharePoint Data Handling

Map semantic, procedural, and episodic data handling across SharePoint and adjacent sources without violating role boundaries.

Admin

Your firm-wide intake and matter status
Permission rules and management exceptions
Reporting, vendor setup, and audit trail

Attorneys

Restricted matter access
Draft review and legal judgment points
Matter history, deadlines, and next actions

Assistants

Broad operational support outside restricted folders
Intake updates and document preparation
Follow-up tasks, scheduling, and status hygiene

Sources can include SharePoint, Outlook, Quo transcripts, Squarespace or website forms, Excel trackers, and approved document templates.

Future operating model

AI gets useful when your firm has three kinds of memory.

Semantic

What your business knows: clients, entities, matters, services, pricing, roles, folders, document types.

Example: who this lead is, what they need, what service applies.

Procedural

How work moves: intake, consult booking, drafting, attorney review, assistant follow-up, status updates.

Example: what happens after a $500 consult is booked.

Episodic

What happened: emails, calls, voicemails, decisions, status changes, drafts, deadlines, and handoffs.

Example: what was said on the latest Mexico lead call.

Three-month roadmap

Month 1: stabilize lead reception.

Month

01

Launch info@ intake response with payment and booking link. Keep scope narrow and measurable.

Pilot Quo for Mexico lead capture. Confirm local number, recording, transcription, and routing.

Define the initial lead and consult data schema. Source, status, owner, consult path, and next action.

Three-month roadmap

Month 2: standardize procedures and role boundaries.

Month

02

Map Admin, Attorney, and Assistant workflows in SharePoint. Respect current permissions from the start.

Document the procedural layer. Intake, consult, draft, review, follow-up, and lifecycle status updates.

Begin template normalization. Editable outputs for client emails, engagement letters, summaries, and slide-style deliverables.

Three-month roadmap

Month 3: make the operating layer compound.

Month

03

Connect website MCP data to discovery and booking. Pricing, services, consult rules, and AI-search visibility.

Use episodic history for better follow-up. Calls, emails, status changes, and decisions become searchable context.

Move from one-off prompts to governed AI workflows. AI can answer what happened, what is next, and what should be generated.

Engagement model

Fractional Chief AI Officer

$5K

per month · three-month plan

Weekly strategy and implementation ownership.

Intake Command Center, Quo pilot, website MCP structure, and SharePoint role/data map.

Confidentiality-first AI architecture and external-tool access model.

Operational documentation so your team can maintain the system after launch.

Prioritized backlog for document generation, dashboards, and future automation.

July

Automate client intake

Launch info@ reception, paid consult response, booking link, and intake queue.

August

Create website MCP

Expose pricing, services, booking rules, and consult paths for AI discovery.

September

Formalize business data model

Map semantic, procedural, and episodic handling for future AI capability.

Upcoming week

This week should turn the proposal into a build plan.

The first week is about confirming constraints, locking the first implementation scope, and giving you room to challenge assumptions before buildout.

01

NDA + access path

Confirm confidentiality terms and whether work begins through a firm email or shared IT inbox.

02

Info@ intake flow

Approve the $500 consult reply, payment link, calendar link, and queue format.

03

Quo pilot checklist

Validate local number, recording, transcription, and routing requirements for Mexico calls.

04

SharePoint role map

Document Admin, Attorney, and Assistant data boundaries before connecting automation.

Questions to address: confidentialityphone providerSharePoint accesstiming and scope