Schedule A Call

Case studies

Practical AI systems, studied in the work.

Workflow case studies and field notes for SME operators who want AI to move real work, with the controls, handoffs, and unglamorous implementation details left in.

Two operations teammates reviewing customer support work together in a naturally lit office

AI EmployeesRole, workflow, owner, access, approval, evaluation, escalation, rollback.

Featured - AI Employees

What Are AI Employees? A Practical Guide for SME Operators

AI employees are governed workflow systems, not magic chatbots. This guide explains what makes them useful inside real SME operations.

7 min readLaunch guide

Case study library

Start with the operating model, then choose the workflow.

These are workflow case studies, not invented client success stories. Named client results stay out until there is real evidence and consent.

A founder and operations consultant mapping a business workflow with paper cards at a table

AuditWorkflow map first. Build second. Retainer only after the system is worth operating.

Why AI Audits Should Come Before AI Implementation

Why SMEs should map workflows, owners, risks, data, and success metrics before buying or building AI automation.

A small operations team organizing, reviewing, and approving routine business documents

WorkflowStart where work is repetitive, measurable, already documented, and safe to review.

The First 10 Workflows SMEs Should Automate With AI

A practical shortlist of repetitive SME workflows that are usually better AI candidates than broad transformation projects.

A paralegal reviewing immigration case documents across organised intake trays in a bright office

WorkflowIntake, classification, extraction, validation queue, human sign-off.

Automating Immigration Case Management: What Actually Works

How document-heavy immigration and legal case workflows are automated in practice, where extraction accuracy really matters, and which steps must stay with a person.

A veterinary receptionist managing appointments while a clinician receives an escalated question

IndustryBooking, reminders, records, paperwork. Never triage.

AI at the Clinic Front Desk: What It Should and Should Not Answer

Practical scope for automating veterinary and clinic front-desk operations, and the clinical line an AI system must never cross.

A technician loading a service van while an office phone and physical booking planner sit unattended

WorkflowEnquiry, qualify, quote band, calendar, confirmation.

From First Message to Confirmed Booking, Without a Human in the Middle

How service businesses automate inbound enquiry to confirmed booking, why response speed decides the outcome, and where the handoff belongs.

A jewellery product photographer creating consistent catalogue images at a professional studio bench

IndustryEmbeddings, index, similarity, stock filter, fallback.

Visual Search for Large Catalogues: How It Actually Works

What image-based product search requires behind the scenes, why zero-result rates matter more than accuracy, and when it is not worth building.

A senior operations expert curating approved manuals at a library-style worktable

AI EmployeesSource approval, citation, refusal, review cadence.

Turning Institutional Knowledge Into an AI Agent Without Losing the Nuance

How to capture the expertise of long-tenured staff into a retrieval system, and why the hard part is deciding what counts as approved knowledge.

A distributor operations specialist validating mixed-format purchase orders against physical parts

OperationsParse, match, validate, exception queue, post.

Eliminating Manual Order Entry: The Unglamorous Automation That Pays

Why re-typing orders from email and PDF into an ERP is the highest-return automation most distributors and manufacturers are not doing.

Editorial rule

No AI hype, no fake case studies, no unsupported metrics.

Every case study must teach a useful operating principle.

Every statistic needs a source or must be framed as a commitment.

Every study should point to one service, one next step, and one related study.

Turn reading into a workflow decision

Learn the model first, or book a call when the pain is obvious.

Join the masterclass Schedule A Call