Start the questionnaire

AI for owner-led businesses

Less admin. Fewer bottlenecks. AI workflows your team can actually use.

We turn repeatable business tasks into tested workflows, organised around your business, with practical lessons and human approval where it matters.

Start the questionnaire Let’s chat

Explore setup options

  • Your business knowledge in one organised workspace.
  • Practical lessons to use, check and run the work.
  • Clear checks, with your team in control of decisions.
A business owner at his desk with a laptop and printed drawings, Table Mountain through the window

Industries that we’ve worked with

  • A travel agent holding up a printed itinerary and pointing at it with a pen over a map, a smiling colleague behind the desk and Table Mountain through the window

    Travel agencies

    • Building itineraries
    • Partner requests and follow-ups
    • Rates and budgets
  • An accountant going through a document with a client across a table with a laptop, a ring binder and a calculator, Table Mountain through the window

    Accounting practices

    • New client onboarding
    • Fee estimates
    • First-draft advice memos
  • A project manager with a tablet talking to a site worker on a building site with scaffolding, Table Mountain and the city behind

    Construction project managers

    • Meeting records
    • Decisions, actions and open questions
    • Comparing contractor pricing
  • A worker checking a tablet in a factory yard while two colleagues load window frames onto a truck, Table Mountain behind

    Manufacturers

    • Turning the calendar into a plan
    • Allocating deliveries to trucks
    • Ordering routes and times
  • Two conservation staff in khaki field shirts at a wooden table on a bush lodge veranda at golden hour, one pointing at a printed report while the other works on a laptop, bushveld and mountains behind

    Non-profits

    • Funder and donor records
    • Partner and collaborator lists
    • Newsletters and impact stories

The starting point

Your business shouldn't live in your head

When every answer, handover and exception comes back to the owner, the team needs more than another chat window.

  1. Too much routes through you

    Make your processes and decision boundaries available to the people doing the work.

  2. The same task starts from scratch

    Give recurring work a clear starting point, trusted inputs and an agreed way to check it.

  3. Handoffs lose context

    Keep the source, the next action and the responsible person together.

  4. New tools don't become habits

    Build confidence through the existing school, practical onboarding and supported adoption.

Work in practice

Where we’ve done this

Examples from client engagements and our own practice, starting with one you can click through.

Client example · Window manufacturing

The calendar is full.
Now plan the trucks and routes.

A window manufacturer supplying carpenters and small installation firms.

The job they wanted off their desk

Turn the delivery calendar into a workable plan for the available drivers. Their routing tool could order one route, but a person still decided which truck took each delivery, in what order and when it left.

That meant weighing rack space, product load, stop limits and road travel times for every truck.

Before

A calendar, a tool that orders one route, and trucks allocated by hand.

Built output

Each truck’s load, stop order and arrival times on real roads, plus a visible list of exceptions.

What we built

AI assistant
Configured to run the agreed tools from a plain-English request and explain status and warnings.
Planning software
Reads a local calendar export, works out each truck’s load, then assigns deliveries to trucks and orders each route on a local road map. Customer addresses stay on the planning computer.
People
Check the detailed workbook locally, resolve exceptions and approve the plan before anyone acts.

Try the workflow

Six fictional orders. Two trucks. Click through to see the routes, and the delivery that still needs a decision.

EXAMPLE WORKFLOWInteractive illustration · fictional data

Follow the work

The operator supplies the delivery date, local calendar export, customer/site files and available drivers. Here we start with six fictional orders.

AI starts the agreed tools; customer details are checked locally by a person.

01 / INPUT

Six orders. One delivery day.

Operator + local tool
Fictional delivery locationsA schematic depot and six delivery stops. Order 06 needs seven rack spaces; each truck holds six.Depot01North Workshop2 racks02Riverside Site3 racks03East Depot1 rack04Hill Workshop2 racks05West Site4 racks06Market Site7 racksFictional delivery locationsA schematic depot and six delivery stops. Order 06 needs seven rack spaces; each truck holds six.Depot01North Workshop2 racks02Riverside Site3 racks03East Depot1 rack04Hill Workshop2 racks05West Site4 racks06Market Site7 racks

Schematic locations · preset example, not a live route calculation

06delivery orders
02available trucks
6 rackscapacity per truck
Inspect the fictional order list

01 · North Workshop2 racks

02 · Riverside Site3 racks

03 · East Depot1 rack

04 · Hill Workshop2 racks

05 · West Site4 racks

06 · Market Site7 racks

Step 1 of 5
Human approval stays in the loop.

Based on implemented planning tools. This website uses fictional orders and preset routes; it runs no AI or routing engine. No delivery system is connected.

The pattern goes beyond deliveries. Start with trusted records, apply the business rules, then give your team a result they can check. We tailor that pattern to other repeatable work.

More examples

Open one to see the task, what we built and the result.

Recorded reporting cyclesTravel business · management reporting
The task
Bring booking and sales-pipeline records into one management view.
What we built
A dashboard and reporting workflow that checks source records, flags missing information and prepares a printable report.
The result
Dashboard refreshes published, with a PDF report exported and archived.
Delivered questionnaireAccounting practice · client intake
The task
Capture the information needed to start a new client engagement.
What we built
An intake questionnaire supported by practical onboarding guides.
The result
A live questionnaire and documented onboarding steps for the team to use.
Shared guide & read-only workbenchAccounting practice · work visibility
The task
Bring work records into a reviewable view and help the team use its workspace.
What we built
A read-only work dashboard based on dated record checks, together with a shared training guide.
The result
A snapshot of work for review and a guide shared with the team.
Output deliveredProperty business · lead preparation
The task
Turn property lead spreadsheets into contact and CRM records.
What we built
A pipeline to research contacts, check for duplicates, update CRM records and track progress in the source spreadsheet.
The result
We operated the workflow and delivered completed spreadsheets and populated CRM records to the client.
Our own practice · internal useConstruction PM · meeting records
The task
Turn a meeting transcript, annotated notes and supporting drawings into a usable record.
What we built
A process for reconciling the sources and organising decisions, actions and unresolved questions.
The result
Draft minutes and an action list prepared for the project manager’s review, with uncertain details flagged.
Our own practice · internal useConstruction PM · cost comparison
The task
Establish what changed between two contractor pricing submissions.
What we built
A line-by-line comparison of quantities, rates, section totals and adjustments.
The result
A reconciliation report showing where rates changed, even though the overall submission total stayed the same.

Your setup includes

Build the workspace. Teach the team.

A personalised business repository and a practical school give your team a shared starting point, then help them put it to work.

Your personalised business repository

Your business, organised for useful work

An organised folder of editable materials: your business context, guidance on trusted sources and instructions for repeatable tasks. We set up your first repository and connect your team to it. Your team then keeps it current and uses it to run the work.

  • Know where the answer comes fromUse authoritative sources and surface gaps.
  • Know what happens nextGive the work a boundary and an accountable person.
  • Know what good looks likeAgree checks and review the result together.

  • Business context

    Your way of working

    Capture your services, voice, people and processes so work starts with the right context.

  • Source guidance

    Use the right records

    Define which sources carry authority and what to do when records are missing, old or contradictory.

  • Practical school & onboarding

    Learn inside your workspace

    Practise business tasks, check the output and learn when to ask for help. The existing school supports the team as it starts using the workspace.

  • Workflow & acceptance

    A useful, bounded result

    Foundation includes two workflows operated by your team, with agreed inputs, outputs and acceptance checks.

How delivery works

Agree the work. Build it. Check it together.

A defined scope makes the result easier to assess and the handover easier to run.

  1. Map

    Choose the workflow, authoritative inputs, responsible people and desired outcome.

  2. Define

    Agree systems, access, approval points, exclusions and acceptance criteria.

  3. Build & check

    Tailor the workspace and test normal work, missing information and the agreed failure cases.

  4. Hand over & adopt

    Give a brief orientation, leave practical lessons and record acceptance. Adoption support stays within the agreed scope.

Our approach

Built from practical experience

We use the same approach in our own day-to-day work: give AI the right context, set clear limits and keep people responsible for decisions.

“We staff our own agency with the agents we build for clients. The operating system you get is the one we run ourselves, on the same platform, every day.”

Don Proudfoot, founder, Build with AI

Three setup options

Choose the scope that fits the work.

Start with a useful workspace, connect defined workflows, or build a more structured operation. Practical lessons, ownership and human review remain part of the approach.

Start here

Your starting point

Foundation

Scoped after discoverySetup price

A personalised business workspace that gives your team the context and confidence to put AI to work, with practical lessons and a tested handover.

Your first repository, set up and connected

2 agreed workflows operated by your team
  • We set up your first business repository and connect your team to it
  • A personalised business workspace with your context and authoritative source guidance
  • Practical lessons through the existing school and onboarding
  • Two agreed, bounded workflows operated by your team
  • Agreed acceptance checks and a tested handover

Final scope and pricing confirmed after your discovery call.

Foundation: typically 14 days to build, 30 days to adopt. Training allowance and adoption support are agreed in your proposal.

Start the questionnaireLet’s chat

Connected workflows

Connected

Scoped after discoverySetup price

Help your team move work between its business tools, with clear checks and approval steps.

4 agreed workflows in total
  • Everything in Foundation
  • Four agreed workflows in total, connected to the supported systems
  • Input/output checks and approval steps
  • Error handling and recovery guidance

Supported systems and integration feasibility are assessed during discovery. We agree delivery dates after discovery. Scope and acceptance criteria are confirmed in your proposal.

Start the questionnaireLet’s chat

Recurring operations

Operations

Scoped after discoverySetup price

Give recurring work a clear queue, visible exceptions and a defined route to approval.

6 agreed workflows in total
  • Everything in Connected, expanded to six agreed workflows in total
  • Scheduling, event-triggered work and queues where feasible
  • Human approvals, exception handling and performance summaries
  • AI may propose and test improvements within agreed boundaries; changes require approval

Feasibility is subject to assessment. We agree delivery dates after discovery. Triggers, monitoring ownership, summary measures and acceptance criteria must be agreed; support for every system is not assumed.

Start the questionnaireLet’s chat

Choose one setup option. Setup fees are confirmed after your discovery call. Your proposal confirms one total setup price for the agreed package, not a fee added to Foundation.

Support after handover

Optional care provides support and maintenance within the scope and hours agreed in your proposal. We quote it separately.

After the 30-day adoption period, it covers agreed updates, fixes and a monthly check-in. Extra development and Operations monitoring are scoped separately.

Your workspace remains usable without a plan. Cancel care with 30 days’ notice.

What does care cover?

Running costs

Paid separately from setup and care:

  • AI and other platform subscriptions
  • API and model usage
  • Hosting and scheduled services
  • Third-party software and integration fees

We agree the providers, account ownership and expected usage before the build.

Prices exclude VAT where applicable. Existing clients retain their agreed scope, price, delivery and support commitments.

Questions

Before you choose

What do we own?
Your tailored workspace and business materials stay yours. The current approach uses Claude Code and your accounts; subscriptions, APIs, platform permissions and connected systems still matter. Documentation and accessible materials help with portability, but moving platforms may require adaptation and testing.
Does the team still need to review the work?
Yes. We define who operates each workflow, what must be checked and where approval is needed. AI can produce errors or miss context. Operations may propose and test improvements within agreed boundaries; it does not make unrestricted changes to the business.
How do you handle sensitive or regulated work?
We scope authorised sources, access limits, approval points, test cases and human review. Data handling and platform suitability must be checked for the specific use. These controls do not guarantee compliance or replace professional judgement. Work requiring professional sign-off remains with the authorised person.
How long does delivery take?
Foundation: typically 14 days to build, 30 days to adopt. We agree delivery dates for Connected and Operations after discovery. Client dependencies, training allowances, support limits and acceptance criteria are confirmed in your proposal. Existing client commitments remain unchanged.
Are the setup fees added together?
No. Setup fees are confirmed after your discovery call. Your proposal confirms one total setup price for the agreed package, not a fee added to Foundation. Optional ongoing care is quoted separately, with scope and support hours confirmed in your proposal. Software subscriptions, model/API usage and hosting are separate from setup and care. Prices exclude VAT where applicable. Upgrade terms and extra work remain to be agreed.
What does ongoing care cover?
Optional ongoing care, quoted separately. Support and maintenance for your agreed workflows. Scope and support hours confirmed in your proposal. Software, API usage and hosting charged separately. Updates, fixes and a monthly check-in follow the 30-day adoption period, within agreed hours. Cancel with 30 days’ notice. A care plan is optional to use your workspace; more complex setups may require a higher care fee. Care does not include unlimited changes. Response times, exclusions and responsibility for operating, hosting and monitoring an Operations build must be agreed separately.
Will business data be shared for measurement?
Basic measurement records remain client-owned. Sharing agreed aggregate diagnostics is optional and separately agreed. The scope confirms which records are needed and whether automated measurement or summaries will be built and verified.

A clear scope from the start

What we agree before starting

We turn your selected package into a written scope, so you know what is included, how it will be checked and who is responsible after handover.

  1. Prices & scopeChoose the package and confirm the total price. Define the workflows, supported systems, integration complexity, users and exclusions.
  2. Delivery & adoptionSet dates, client dependencies, training allowances and the adoption-support limits for each tier.
  3. AcceptanceAgree evidence, thresholds and sign-off for normal runs, missing inputs, approvals, errors, recovery, queues and schedules.
  4. Care & ongoing responsibilitySet included hours and response times. Name who operates, hosts and monitors Operations after handover.
  5. Terms & measurementAgree payment milestones, changes, upgrades, extra work, measurement records and any optional aggregate sharing.
  6. Accounts & running costsAgree the client-owned accounts, permissions, subscriptions, API usage, hosting and third-party software needed for the work.

Start with the work.

Email Don with a repeatable task that slows your team down. We’ll discuss a useful first workflow and a scope you can assess before you commit.

Start the questionnaire Let’s chat

don@builtwithai.co.za