Years of client files, notes and statements, and still nobody can find an answer without opening ten documents?
What we do: We ground AI in your own records, so every answer comes from your files, with the source attached for the adviser to check.
Scaling the companies that scale the economy.
Log / visit / build / hand over
Packages
Five packages. Each one does one job, start to finish.
Every package runs in three phases with one engineer at your firm.
-
01Most requested
Agent Security
Your client data protected from AI misuse, inside and out.
From R265,000 · 3 phases, fixed
What’s included → -
02Most requested
Process Automation
One internal process, redesigned and automated end to end.
R195,000 · 3 phases, fixed
What’s included → -
03
FSP Back Office
The admin behind every client file, done by agents your team supervises.
From R285,000 · 3 phases, fixed
What’s included → -
04
Client Work Agents
One agent that does one piece of client work properly.
R295,000 · 3 phases, fixed
What’s included → -
05
Investment Models
A model built on your investment philosophy, that you can explain one trade at a time.
From R1,250,000 · 3 phases, fixed
What’s included →
Sound familiar?
If any of these sound like your practice, you’re in the right place.
What FSP owners and managers tell us on the first call, in their own words.
Platforms and suppliers hold your clients’ data, but you couldn’t say how exposed you are if one of them is breached?
What we do: We map who holds your client data, test your systems the way an attacker would, and give you a plain list of what to fix first.
Your staff already paste client details into AI tools, and nobody is sure what those tools keep?
What we do: Guardrails before capability: approved tools, scoped access, masked personal data and a person signing off, in line with POPIA.
Records of advice, FICA packs and review packs eat your advisers’ week?
What we do: Agents draft them from the meeting and your platforms, and an adviser approves every one. That is our FSP Back Office package.
Your administrators type the same client into three different platforms?
What we do: Agents move the data and fill in the forms, so your people get back to clients and the work you hired them for.
You’ve tried ChatGPT and a few AI tools, but none of it changed how your practice actually runs?
What we do: We build AI into the platforms you already use and train your team on it, so it becomes part of the daily work.
Your AI bill keeps growing, but you can’t point to what it’s saving you?
What we do: A handful of workflows taken to production, the right model for each task, and cost tracked against the time it saves.
Your automations fall over every time a form, a supplier or a screen changes?
What we do: Agents that work from the goal instead of a fixed script, so they adapt when the interface does.
How we operate
Three phases. One path from problem to solution.
Every package is the same three phases: develop, implement beside your team, and hand over. Before them comes Phase 0, information gathering, where we find the problem (the first 10 hours are on us). After them comes an optional monthly run service. Cybersecurity runs through all of it: checked in Phase 0, built in during the build, drilled at handover, and watched for if you keep us on.
0
Phase 0: information gathering
A first call, your confidential log, a visit to your office, an assessment, and written findings with a fixed price.
Problem scoping
Bring the problem as you feel it. We map where the time, money and risk actually leak.
We sit with the people doing the work, walk through the last time it went wrong, and trace the job start to finish. You do not need a brief. We turn the frustration into a written statement of the problem, the data you already hold, and what a fix would have to do to be worth it.
Security posture check
Before we build anything, we look at what you are exposed to right now.
A threat model of the systems the new work will touch, a check of what is publicly visible about them (only with your written permission), and a POPIA readiness check on the data involved. Anything we build has to make you safer, not add a new door. This also surfaces the quick security wins most businesses do not know they have.
Findings and a fixed price
Written findings you keep whether or not you go on, and one package at one fixed price.
What each problem costs you, what has to be in place first, which of our five packages fits, and the acceptance test we would be held to. Written so your board or bank manager can read it. It is yours to take anywhere.
What you get
- A written problem statement in plain language, not a brief you had to produce yourself
- A security posture check on every system the work will touch
- A phased plan you own outright, usable with us or anyone else
- A clear yes/no on whether the fix is worth building
How Phase 0 works. It starts with a first call. Then you log your struggles privately on our confidential form: an automatic email with your next steps arrives a minute later, and the people assigned to your firm read the log itself within one working day. Then a visit: someone from our team sits with the people doing the work and logs each problem as it happens. Last come the assessment and your written findings. The first 10 hours carry no invoice and no obligation. If there is nothing worth building, we say so and part as friends.
1–2
Phases 1 and 2: develop and implement
Phase one, one engineer builds it on your accounts. Phase two, it runs beside your team until it is right.
Production build
Working software demonstrated every week. No slide decks, no reveal at the end.
Agents, automation pipelines and data tooling built on your own hosting, database and AI accounts from day one. Each week ends with a demonstration of live software and sets the next week's priorities. Scope is locked before the final week.
Secure by construction
Guardrails, scoped permissions, masked personal data and a human sign-off, built in rather than bolted on.
Every system ships with least-privilege access, an audit log of what it did and what a person changed, ID numbers and bank details masked before any text reaches a model, and POPIA transfer terms documented.
Testing with your people
Your staff use it before you decide anything. Their change log drives the final version.
Real jobs, real data, a log of every edit they had to make. Held-back examples are regenerated and compared with the originals. Voice, structure and layout corrections are applied as they are found, so what you sign off is what they will actually use.
What you get
- Working software you can see and use every week of the build, not just at the end
- A system secured by design: least-privilege access, masked personal data, a full audit log
- A version your own staff have already tested against real jobs and real data
- Everything built on your own hosting, database and AI accounts from day one
3
Phase 3: hand over
Your staff are trained, the acceptance test is run with you, the keys are handed over, then we step back.
Staff training
The people who do the job learn to run it, change it, and know when to worry.
Hands-on sessions on the live system, not a manual left on a desk. Operators learn the daily routine, managers learn the reports, and one named person learns how to adjust settings and re-run the learning passes.
Security awareness & incident drill
Staff learn what a phishing attempt, a leaked credential or a misbehaving agent looks like, and what to do next.
A short awareness session for everyone who touches the system, a written incident playbook with the phone numbers on it, and one tabletop drill so the first time your team handles an incident is not a real one.
Keys, code and documentation
Source code, hosting, database and AI credentials handed over in your company's name.
Nothing sits in a MAKGLOBAL account. You get the repository, the infrastructure logins, written documentation and a final progress report. Two of your people are trained to run it, so nothing depends on our engineer after handover, and there is no lock-in.
What you get
- Staff who can run the system end to end without us in the room
- A written incident playbook and one completed tabletop drill
- Every credential, the source code and documentation, in your company's name
- The acceptance test run with you; the last 30% is only invoiced when it passes
+
Optional run, month to month
We keep it working and secure. Your team stays in charge, and you can stop on 30 days’ notice.
Monitoring and fixes
We watch the system and fix what breaks when a platform, form or model changes.
Models get retired and your templates change. We switch in better models without a rebuild, adjust templates, cover the model and API costs agreed in the package, and meet you once a month to review what the system did.
Quarterly security review
A retest of the controls we built, and a review of any new AI tool before it touches client data.
Every quarter we retest access, logging and approvals, review new tools and suppliers your team wants to use, and update your AI use policy. Between reviews, automated monitoring watches your agents around the clock and every alert reaches our team.
The next package
Once the first system is running, we look at the next job together.
The next package starts with a short visit again, and goes faster because we already know your firm, your data and your people.
What it costs
- From R9,500 to R35,000 a month, depending on the package (shown on each package)
- Month to month, cancellable on 30 days’ notice, no lock-in (Investment Models: 12-month minimum)
- Never required: handover is built so your team runs it without us
The path, side by side
| Phase 0 | Phases 1–2 | Phase 3 | Optional run | |
|---|---|---|---|---|
| What it is | We scope the problem, check your security posture, and write the plan. | We build the plan into working, secured software on your own accounts. | Your staff are trained, the keys and code are handed over, we step back. | Optional monthly care: monitoring, fixes, model updates and a quarterly security review. |
| Speed | 10 free hours, findings within five working days | Phases one and two | Phase three | Month to month on 30 days’ notice (Investment Models: 12-month minimum) |
| Output | Confidential log, visit notes, security posture check, written findings and a fixed price. | Production system, audit log, masked personal data, weekly demos, written progress reports. | Trained staff, incident playbook, source code, credentials and documentation in your name. | Monthly review, model and template updates, quarterly security retest, first-line support. |
| Who it is for | The owner or MD with a nagging problem and no brief. The entry point. | Owners and operations leads who have signed off the plan. | The staff who will run it every day, plus one named administrator. | Firms that want it looked after without hiring for it. |
Why MAKGLOBAL
A dedicated team, around the clock, at the heart of the problem.
- A dedicated team. The same engineer and lead from the first call to handover, who learn your clients, your platforms and your people.
- Around the clock. Your agents work day and night, and our monitoring watches them day and night, so problems are caught before your clients notice.
- The heart of the problem. We sit with the people doing the work and fix the cause, not the symptom. No plasters.
Talk to us
Book a 30-minute first call.
Five quick answers so we only take calls we can help with, then pick a time from Tuesday to Thursday. If no time suits, ask for a call back in the same form.
- No invoice and no obligation for the 10 hours
- You keep the written findings either way
See open times
Compare the packages first
Or email hello@makglobal.africa. A person reads every message.
Common questions
Questions people ask
Are the 10 hours really free?
Yes. The first 10 hours of our time come with no invoice and no obligation to continue. That covers the conversations, the site visit and the digging into your data and systems. If there is nothing worth building, we say so and part as friends.
Do we have to start with Phase 0?
Almost everyone does, because it is where the plan comes from. If you already have a scoped project, we can compress Phase 0 to a short review of your plan and start Phase 1 sooner.
Where does cybersecurity fit?
Everywhere. Phase 0 includes a security posture check on the systems the work will touch. Phases 1 and 2 build in least-privilege access, masked personal data and an audit log. Phase 3 (handover) includes staff awareness training and an incident drill. The optional run fee includes a quarterly security review. Agent Security is also a package on its own.
Who owns what we build?
You do, outright. Hosting, database and AI accounts are opened in your company's name at the start of Phase 1. Source code lives in your repository. At handover you have every credential and the documentation. There is no lock-in.
What if the scope grows during the build?
Scope is locked before the end of Phase 2. Anything outside it is written up as an extension before any work starts. You decide.
How long does a package take?
Every package runs in three phases: develop, implement and hand over. We agree the dates for each phase with you before Phase 1 starts, put them in the scope sheet you sign, and hold ourselves to them. You see working software at every phase, not just at the end.
What does it cost?
Each package has one fixed price, from R195,000 excl. VAT, shown on the packages page. You pay 40% on signature, 30% at go-live (end of phase two), and the last 30% only when the acceptance test agreed before Phase 1 passes.
Do we have to keep paying after handover?
No. Handover is designed so your team runs the system without us. The monthly run fee is for firms that would rather have us monitor, fix and review it than hire for it. It is month to month and cancellable on 30 days' notice, except Investment Models, which has a 12-month minimum.