Systems Architecture Consultation

About This Document

This document is the output of a structured systems architecture consultation. It maps Alex's current operating state, identifies gaps and risks, designs a coherent system architecture, classifies every key function by who or what should handle it, and sequences the recommendations into a prioritised roadmap.

It is not a tools list. Every recommendation here is grounded in how Alex actually works — the goal is systems that fit the founder, not systems the founder has to accommodate.

Client Brief

Who

Alex is an executive coach and leadership consultant with 15 years of corporate L&D and organisational design experience. 18 months into running independently, Alex works with senior leaders in high-growth technology companies.

Current offering: intensive 1:1 coaching engagements (6–12 months), occasional team workshops, and a nascent group programme in development. Revenue comes primarily from 1:1 work.

The Problem as Presented

"I'm fully booked. I can't take on more 1:1 clients, but I also can't seem to get the group programme built. Everything admin-related falls to me and it takes about 8–10 hours a week I can't afford to lose."

Alex came looking for workflow optimisation. What the audit revealed was different: the constraint isn't time — it's that the business has no architecture. Everything flows through Alex personally because there are no systems to route it elsewhere.

Engagement Scope

Current state audit across all business domains

System architecture design (integrated, not siloed)

Allocated intelligence classification (human / AI / automated)

SOP requirements identification

Sequenced build roadmap

What the Audit Found

The core delivery (coaching) is excellent. Everything around it is ad hoc.

Knowledge and IP lives in Alex's head, not in a system — making it impossible to scale, delegate, or build from.

Eight to ten hours of weekly admin is largely rule-based work being done manually. Most of it can be removed from Alex's plate immediately.

The group programme isn't blocked by time — it's blocked by the absence of a knowledge capture system and a content production workflow.

There is no pipeline visibility. Alex doesn't know where leads are, which convert, or what the revenue trajectory looks like.

Engagement Summary

Domains at Risk

Functional Gaps

Domains Assessed

Functions Mapped

SOPs Required

Current State Assessment

Each domain has been assessed against four maturity levels: 1 — Not Started (function is absent), 2 — Functional (works but manual, fragile, or disconnected), 3 — Optimised (reliable and efficient for current scale), 4 — Integrated (connected to other systems, data flows, compound value). Most domains sit at Level 2 — functional enough to run the business today, but not designed to scale.

Knowledge & Intelligence

Research, IP, client insights, and the connections between them

Level 1 — Not Started

What's Working

Alex has deep domain expertise, well-developed frameworks, and genuine intellectual rigour

Client session notes are kept (inconsistently) in Notion

Strong intuitive pattern recognition across clients

At Risk

All core IP and frameworks exist only in Alex's head — no structured documentation

Session notes are inconsistent and not connected to anything

Research is scattered across bookmarks, emails, and physical notes

If capacity drops, nothing can be delegated — there's no context to hand over

The Gap

No central knowledge system — frameworks, research, and client patterns have nowhere to live

No capture habit or process — ideas disappear

No connection between client work and product development

Level 1 / 4

Create & Communicate

Content, thought leadership, and platform presence

Level 2 — Functional

What's Working

LinkedIn presence is active and generates inbound interest

A newsletter (Mailchimp) goes out monthly — audience is engaged when it arrives

Alex writes well and has a distinctive voice

At Risk

Content creation is slow (2–3 hours per piece) because it starts from scratch each time

No connection between what Alex is researching and what gets published

Inconsistent posting — energy-dependent, not system-supported

Newsletter and LinkedIn operate as separate, unrelated streams

The Gap

No content workflow — each piece is a fresh effort

No repurposing system — ideas don't compound across channels

No brand asset library — visual production is ad hoc

Level 2 / 4

Attract & Convert

Lead generation, pipeline, and the path from interest to client

Level 2 — Functional

What's Working

Calendly is in use for discovery call scheduling — this works well

Referrals are strong — most clients come via word of mouth

Conversion rate from discovery call to proposal is high

At Risk

No pipeline visibility — leads are managed via email and memory

Proposals are Word documents created manually each time (1–2 hours each)

Follow-up is inconsistent — people fall through the cracks

No onboarding system — new client setup is a manual flurry each time

The Gap

No CRM — pipeline is invisible

No proposal templates — manual creation wastes time on a repeatable task

No automated onboarding — first impression is inconsistent

Level 2 / 4

Deliver

Coaching, programmes, and everything the client experiences

Level 3 — Optimised

What's Working

1:1 coaching quality is high — outcomes are strong, referrals reflect this

Session structure is developed and consistent

Alex's methodology is well-formed, even if not yet documented

At Risk

Session prep requires pulling from scattered sources (email, Notion, memory)

Post-session summaries are inconsistent — some clients get them, some don't

Group programme is blocked — no content system, no workflow, no structure to build into

The Gap

Methodology undocumented — can't be productised, delegated, or systematised

No client portal or delivery system for scalable programmes

No post-session workflow — insights aren't captured systematically

Level 3 / 4

Sustain

Finance, admin, and the operational layer that keeps the business healthy

Level 2 — Functional

What's Working

Xero is set up and accounting is clean

Invoices are paid reliably — no outstanding debts

Alex has good intuition about where the business is financially

At Risk

Invoice generation is fully manual — triggered by memory, not by system milestone

8–10 hours of weekly admin is almost entirely rule-based work done by hand

No analytics or business intelligence — no forward visibility

Processes are undocumented — nothing can be delegated or automated reliably

The Gap

No billing automation — invoices require manual creation and sending

No documented processes — automation requires SOPs first

No dashboard — business health requires manual checking

Level 2 / 4

System Architecture

This map shows how the business needs to be structured — not just what tools exist, but how each domain connects to the others and what sits at the centre. The maturity levels from the Current State assessment are reflected here: domains shown in amber or red are functional but fragile, and their health affects everything connected to them.

The Knowledge system is deliberately positioned as the central nervous system. This is an architectural decision: when research, client insights, frameworks, and content ideas all flow through a single connected system, each domain benefits from the others. Without this, everything remains in silos.

System Map — Target Architecture

Build First

Knowledge & Intelligence

Central Knowledge System

IP & Framework Library

Client Insight Capture

Research Organisation

Optimise

Create & Communicate

Content Production Workflow

Newsletter (Mailchimp)

LinkedIn & Platform Strategy

Brand Asset Library

Automate

Sustain

Billing & Finance (Xero)

Invoice Automation

Business Analytics

Process Documentation

Critical Gap

Attract & Convert

Scheduling (Calendly)

CRM & Pipeline

Proposal Templates

Client Onboarding

Extend

Deliver

1:1 Coaching (strong)

Session Prep & Notes Workflow

Group Programme

Client Portal / Delivery System

Not started / critical gap

Functional — needs optimisation

Optimised — maintain & extend

Integrated — target state

Key information flows: Client session insights → Knowledge System → Content production. Research → Knowledge System → IP Library → Group Programme. CRM data → Pipeline visibility → Revenue forecasting. Every domain feeds back into Knowledge — this is what creates compound value over time rather than isolated effort.

What Is Allocated Intelligence?

Every task in your business requires a certain type of intelligence to complete it. Some require your specific expertise, relationships, and judgment — no system can substitute. Others require consistent execution of a known process — these are the tasks stealing your capacity. The goal of allocated intelligence is to ensure the right type of intelligence is doing each job.

Below, every key function in Alex's business is classified across four categories. Where tasks are allocated to agents or automations, the required SOPs are flagged — because intelligent allocation only works when the system has sufficient context to operate reliably.

Human-Led

Requires Alex's specific expertise, relationships, or judgment. Cannot and should not be delegated.

AI-Assisted

Alex leads; AI accelerates and augments. Human in the loop for quality and direction.

Agent-Handled

AI can own this with sufficient context and clear SOPs. Requires setup, not ongoing attention.

Automated

Rules-based, no intelligence required. Triggered by events. Should not touch Alex's time at all.

Knowledge & Intelligence

Function / Task	Allocation	Rationale	SOP Required	SOP Status
Framework and IP development	Human-Led	Alex's core intellectual work. The expertise is the product.	—	—
Research synthesis and sense-making	AI-Assisted	Alex decides what's relevant; AI synthesises, summarises, and surfaces connections across sources.	Research capture template	Needed
Session note capture and tagging	Agent-Handled	Structured capture from a consistent template can be handled by AI. Alex reviews, doesn't produce.	Session notes template, tagging taxonomy	Needed
IP documentation (structuring frameworks)	AI-Assisted	Alex provides the thinking; AI provides the structure, formatting, and drafting scaffolding.	IP documentation format guide	Needed
Pattern identification across clients	AI-Assisted	With structured session notes feeding a knowledge system, AI can surface emerging patterns. Alex interprets.	Session notes template (above)	Needed

Create & Communicate

Function / Task	Allocation	Rationale	SOP Required	SOP Status
Content strategy and platform decisions	Human-Led	Strategic choices about where to show up and why require Alex's judgment and positioning decisions.	—	—
Long-form articles and newsletter writing	AI-Assisted	Alex provides the ideas, angles, and insight; AI drafts, structures, and refines. Voice must be Alex's.	Voice & tone guide, content pillars	Needed
Social posts from existing content	Agent-Handled	Repurposing an approved piece into LinkedIn/newsletter formats is consistent enough for agent execution.	Voice & tone guide, format templates per platform	Needed
Visual asset creation (headers, social graphics)	Agent-Handled	With a brand guide and templates, visual production becomes a consistent, repeatable task.	Brand asset library, visual templates	Needed
Content scheduling and publishing	Automated	Once content is approved, scheduling is purely logistical. Tools handle this entirely.	—	—

Attract & Convert

Function / Task	Allocation	Rationale	SOP Required	SOP Status
Relationship building and networking	Human-Led	Trust and referrals are Alex's primary channel. This is irreplaceable human work.	—	—
Discovery call conversations	Human-Led	Initial qualification and fit assessment requires Alex's judgment. This is also a first impression.	—	—
Initial lead qualification (pre-call)	Agent-Handled	An intake form with structured questions can be reviewed by AI to assess fit before Alex's time is committed.	Ideal client profile, qualification criteria	Needed
Proposal creation	AI-Assisted	Core structure and language is repeatable. AI drafts from a template; Alex customises the positioning.	Proposal template, service descriptions, pricing framework	Needed
Follow-up sequences (post-discovery)	Automated	Timed, triggered follow-up sequences require no intelligence — just consistency. Currently done by hand.	—	—
Client onboarding (new client setup)	Agent-Handled	Welcome, contracts, intake, scheduling setup — all of this is consistent process. Currently takes Alex 2+ hours per client.	Onboarding checklist, welcome sequence, contract template	Needed
Discovery call scheduling	Automated	Already handled by Calendly. Working well — extend to include pre-call intake form.	—	In place

Deliver

Function / Task	Allocation	Rationale	SOP Required	SOP Status
1:1 coaching sessions	Human-Led	This is the core product. Alex's expertise and presence are the value.	—	—
Session preparation (notes, agenda)	AI-Assisted	AI pulls previous session notes, surfaces themes, and drafts a prep brief. Alex reviews and adds focus.	Session notes template (above)	Needed
Post-session notes and action capture	Agent-Handled	With a voice memo or transcript input, AI can produce structured session notes. Alex reviews before sending.	Session summary format, client context file	Needed
Group programme design	AI-Assisted	Curriculum architecture and content structure can be scaffolded with AI; Alex provides the IP and methodology.	IP library (once documented)	Needed
Client progress tracking	Automated	Milestone tracking in a simple CRM or project tool — triggered by session completion.	—	—

Sustain

Function / Task	Allocation	Rationale	SOP Required	SOP Status
Financial strategy and decisions	Human-Led	Pricing, investment decisions, and business model choices require Alex's judgment.	—	—
Invoice generation and sending	Automated	Triggered by project milestone or date — Xero can handle this end-to-end with no Alex involvement.	Invoice trigger rules (milestone map)	Needed
Expense categorisation and tracking	Agent-Handled	With clear categorisation rules, AI can process receipts and classify expenses. Alex reviews monthly.	Expense categorisation rules	Needed
Monthly financial summary	Agent-Handled	AI aggregates and presents; Alex interprets. This should arrive in inbox, not require active creation.	Financial summary template	Needed
Process documentation	AI-Assisted	Alex describes the process; AI structures and documents. This is how the SOPs below get created.	—	—

Required SOP Inventory

Why SOPs Come First

Intelligent allocation — agents, automations, AI-assisted workflows — only works when the system has enough context and instruction to operate reliably. An AI asked to qualify a lead without an ideal client profile will produce inconsistent results. An agent producing session notes without a template will produce noise. The SOPs below are not bureaucracy; they are the context that makes allocated intelligence functional. Building these is Phase 1.

Voice & Tone Guide

Alex's writing voice, tone principles, what to avoid, and examples of on-brand vs off-brand content. Required before any AI-assisted or agent-handled content production.

Session Notes Template

Structured format for capturing session content — key themes, commitments, observations, progress markers. Enables AI-assisted prep and post-session workflows.

Ideal Client Profile

Specific criteria for a strong-fit client — sector, role, stage, readiness indicators, and red flags. Required for lead qualification agent to work reliably.

Client Onboarding Checklist

Step-by-step process for welcoming and setting up a new client — contracts, intake, scheduling, welcome communication. Allows this to be agent-handled from day one.

Proposal Template & Service Descriptions

Core proposal structure, service option descriptions, and pricing framework. Reduces proposal creation from 2 hours to 20 minutes with AI assistance.

Content Pillars Document

Alex's 4–5 core content themes, what each covers, and how they connect to the methodology. Gives AI-assisted content production a consistent frame to work within.

Invoice Trigger Rules

Map of when invoices are triggered per service type — at which milestone, what amount, what payment terms. Enables full billing automation via Xero.

IP & Framework Documentation Format

Template and process for extracting Alex's tacit knowledge into structured, reusable IP documents. Required before the group programme can be designed or delivered at scale.

Discovery Call Scheduling (Calendly)

Already in place and working. Recommend extending with a pre-call intake form to enable lead qualification agent.

Build Sequence & Recommendations

The sequence below is not arbitrary. Each phase unlocks the next: SOPs enable agents, agents free capacity, freed capacity builds the knowledge system, the knowledge system enables the group programme. Skipping phases creates fragile systems that depend on Alex's manual oversight to function — which defeats the purpose.

The immediate priority is not building new systems. It is documenting what already exists so that intelligent allocation becomes possible. Most of the 8–10 hours of weekly admin can be removed from Alex's plate within 4–6 weeks with this foundation in place.

Phase 1 — Now

Build the Foundation

Weeks 1–4 · SOPs, knowledge system, quick automation wins

Document the SOPs

Nothing can be automated or agent-handled without process documentation. Start with Voice & Tone Guide, Session Notes Template, and Onboarding Checklist — these unlock the most immediate capacity.

Expected Outcome2–3 hours per week recovered immediately from client onboarding alone.

Set Up Central Knowledge System

The most important architectural decision in the whole system. Everything else connects here. Doesn't need to be complex — it needs to be consistent. Start with a clear capture habit and tagging structure.

Expected OutcomeIdeas, research, and client insights stop disappearing. The group programme starts to become buildable.

Automate Invoicing (Xero)

One of the cleanest automation wins in the whole system. Define the milestone triggers, configure Xero, and remove this from Alex's weekly task list permanently.

Expected OutcomeInvoicing becomes invisible. 1–2 hours per month recovered.

Build Proposal Templates

High-effort, repeatable work with a strong, consistent pattern. Proposal creation drops from 2 hours to 20 minutes once templates are in place and AI is assisting the customisation.

Expected OutcomeProposals feel more professional and take a fraction of the time.

Phase 2 — Next

Activate Allocated Intelligence

Weeks 5–10 · CRM, content workflow, session note system

Implement CRM & Pipeline Visibility

Alex currently has no forward revenue visibility. A lightweight CRM — even a well-structured Notion database — immediately shows where leads are, what's converting, and what follow-up is needed. The data foundation for future analytics.

Expected OutcomeLead visibility. Follow-up stops being memory-dependent. Revenue predictability improves.

Build Content Production Workflow

Content creation is currently starting from scratch every time. With the Knowledge System in place and a workflow connecting research → draft → repurpose, content volume can increase without a corresponding increase in time.

Expected OutcomeContent time halved. Publishing consistency improved. LinkedIn and newsletter no longer feel like separate efforts.

Activate Session Notes Agent

Once the Session Notes Template SOP exists, AI can take a voice memo or transcript and produce structured notes in seconds. This closes the loop between delivery and knowledge — every session feeds the system.

Expected OutcomePost-session admin drops to near zero. Client insights start accumulating as structured data.

Configure Follow-up Automation

Post-discovery follow-up is currently done by hand when Alex remembers to do it. Automated sequences triggered by discovery call completion ensure no lead falls through the gap without Alex's attention.

Expected OutcomeConversion rates improve. Follow-up is no longer a source of anxiety or manual effort.

Phase 3 — Growing

Build the Scalable Product

Months 3–5 · Group programme, delivery system, refinement

Document IP & Design Group Programme

With a functioning knowledge system and the IP documentation SOP, Alex now has the infrastructure to externalise the methodology. The group programme becomes buildable because the raw material is no longer locked inside a single person's head.

Expected OutcomeGroup programme moves from "in development" to "in delivery." Second revenue stream activated.

Set Up Delivery System for Programmes

Scalable group delivery needs infrastructure: a client-facing portal, session materials, progress tracking, and communication that doesn't run through Alex's inbox. Design this to fit the programme, not the other way around.

Expected OutcomeGroup programme can run without consuming 1:1-equivalent time. Delivery becomes repeatable.

Refine Allocated Intelligence Layer

At this stage, there's real operational data to work with. Review which allocations are working, which agents are producing reliable output, and where SOPs need tightening. Systems improve most when there's feedback from actual use.

Expected OutcomeWeekly admin time drops to under 3 hours. Capacity for 1:1 work increases or can be partially replaced by group revenue.

Phase 4 — Mature

Intelligence & Compounding

Months 6+ · Analytics, feedback loops, strategic intelligence

Build Business Analytics Dashboard

Once data is flowing through a connected system — CRM, delivery, finance, content — an analytics layer becomes genuinely meaningful. A weekly business pulse report surfaces patterns that aren't visible when managing each system separately.

Expected OutcomeStrategic decisions are data-informed. Resource allocation follows evidence rather than instinct.

Activate Client Pattern Intelligence

With structured session notes accumulating across clients over time, AI can surface recurring themes, common obstacles, and emerging patterns in the client base. This feeds both product development and content strategy — closing the loop between delivery and growth.

Expected OutcomeDelivery makes the business smarter. Every client engagement contributes to IP development and programme improvement.

A Note on Sequence

The temptation in systems design is to start with the most interesting or visible problem — in Alex's case, probably the group programme or the CRM. The architecture above says: start with the invisible infrastructure first.

SOPs before agents. Knowledge system before content workflow. Foundation before product. This sequencing feels slower in week one and dramatically faster by month three. Systems built on a solid foundation adapt and improve. Systems built reactively require constant manual intervention to keep running — which is exactly where Alex started.

The goal is not a perfect system. The goal is a system designed around how Alex actually works — one that compounds over time, adapts as the business grows, and returns hours to the work that only Alex can do.