Put AI to work in your business. Live in weeks, not years.
We place a MyBusiness App engineer inside your team. They learn how the work really gets done, find the spreadsheets, inboxes and manual handoffs slowing you down, and build AI agents that run on your own data and systems. The same way we run MyPetParlor, MyBeautySalon, MyBillingHub and MyMicroFinance.
What Custom AI Agents delivers
- Routine work like intake, triage, follow-ups and reconciliations handled by AI agents, with your staff in control
- WhatsApp and email assistants that answer your customers from your own records, day and night
- Reports that used to take a week, ready when you ask for them, straight from your own systems
Built for teams like yours
Operations teams
Back offices buried in manual capture, document checks, follow-ups and status chasing.
Product and IT leaders
Teams that need AI features in production but cannot hire and ramp an AI team in time.
Regulated businesses
Financial services, insurance, healthcare and logistics, where every automated decision needs an audit trail.
Three steps, with evidence at every one
Each step ends with something you can inspect, so you decide whether to continue on results rather than promises.
- 12 weeks
Define
Your engineer works alongside the people doing the job, maps the data and rules involved, and agrees with you what "good" looks like before anything is built.
- Vision and scope document
- Data and systems map
- Acceptance criteria signed off by your specialists
- 24–6 weeks
Build and prove
A working agent is built against real cases and tested against answers your specialists have already validated, so you see evidence before you commit to production.
- Working agent on your infrastructure or ours
- Test results against validated cases
- Weekly demos to the business owner
- 3Ongoing
Launch and improve
The agent goes live with monitoring, human review where it matters, and a clear owner. Your engineer keeps improving it and moves on to the next workflow.
- Production rollout with monitoring
- Operating guide for the team
- Monthly quality review
Clear prices to start with
Prices in South African rand, excluding VAT. International engagements are quoted in USD or GBP on request.
Discovery
R 45 000
2 weeks, fixed price
Find out which workflow is worth automating first and exactly what it would take.
- On-site or remote discovery with your team
- Prioritised list of automation opportunities
- Scope, data map and acceptance criteria
- Fixed quote for the build
Pilot agent
Most chosenFrom R 180 000
4–6 weeks
One agent, built and proven on real cases, ready to switch on in production.
- Everything in Discovery
- One production-ready agent
- Evaluation against validated cases
- Handover and operating guide
Embedded AI engineer
R 135 000 / month
Month to month
A dedicated forward-deployed engineer working inside your team, building agent after agent.
- Full-time senior AI engineer
- Continuous build, launch and improvement
- Monthly outcomes report to leadership
- Cancel with 30 days notice
Custom AI Agents questions
What enterprises usually ask before engaging us.
What kind of work is a good fit for an AI agent?
Work that is repetitive, rules-driven and currently done by reading documents, copying between systems or chasing people. If your specialists can explain how they decide, we can usually build an agent that does the first pass and hands them the exceptions.
Does the engineer really work on site?
Yes, where it helps. Most engagements mix time on site with your team and remote build time. The point is that the engineer learns the work from the people who do it, rather than from a requirements document.
What happens to our data?
Your data stays in your systems or in a tenant dedicated to you. We sign an NDA before discovery, work within your POPIA and security policies, and never use your data to train shared models.
Who owns what gets built?
You do. The code, prompts, evaluations and documentation are handed over, and you can run the agents yourselves or keep us on to operate and improve them.
What if the agent makes a mistake?
Every agent is tested against cases your specialists have checked before it goes live, and decisions that carry risk are routed to a person for approval. Monitoring flags drift so it is caught early.
Talk to us about Custom AI Agents
Tell us a little about your team and what you want to achieve. We reply within one business day, and can respond to tenders and RFPs.