What is the AI marketplace?
The AI marketplace is a catalogue of ready-to-ship AI employees a South African business can browse, search and put to work: WhatsApp agents, inbound and outbound AI callers, website chat, and back-office automations. Every listing in the AI marketplace is a working build, not a slide or a demo video. The channel, the tools the agent may call, and the point where a human takes over are defined before anything goes live.
The AI marketplace exists because most companies do not need a research project. They need a specific job covered, starting this month. So the catalogue is organised by job, not by model: reception, lead qualification, quote follow-up, booking, document handling, reporting. We run the AI marketplace from Cape Town and have delivered systems of this kind for 35+ companies over 3+ years, 340+ solutions built. If you do not adopt, competitors will. If you do, we make it pay for itself.
How does hiring an AI employee work in practice?
Hiring an AI employee works like onboarding a junior team member who arrives with a written job description. Scope, channel and escalation rules are agreed first. The agent is then grounded in company policy, price lists and documents, so answers come from the business rather than from guesswork. Only then does the agent get keys to the tools it is allowed to read and write.
On shift, the agent reads context, calls those tools, updates the CRM record, and hands off to a person with the full transcript attached. Nothing arrives at a colleague as a cold ping. Wording is drafted, reviewed and approved before a single message sends, and every action lands in an audit trail. SLA timers, alerts and weekly changelogs show what the AI employee did, where it escalated, and what changed since the previous week. We assemble the steps with n8n or Make.com, with language handled by OpenAI, Anthropic Claude or Google Gemini.
Which AI employees can a South African business hire?
South African businesses can hire four broad categories of AI employee from the AI marketplace. WhatsApp agents answer enquiries, qualify leads and book slots on the channel most local customers already open. AI callers cover reception, bookings, follow-ups and quote chasing, with call summaries pushed straight to WhatsApp and the CRM updated in real time. Website chat catches the visitor who was about to leave.
Back-office automations sit behind all of that, moving data between ERPs, spreadsheets, accounting systems and custom apps so nobody retypes anything. Agentic orchestration ties the group together: agents that read context, call tools, update systems, and hand off to humans with transcripts. Analytics run across ads, chat, calls and pipeline, so the business sees which AI employee is earning its place. Roles can be hired one at a time. Most companies start with a single agent and add the next once the first is trusted.
Does the AI marketplace work with our existing tools?
The AI marketplace is built to plug into the stack a company already runs, not to replace it. Integration is the core of the work, not an add-on at the end. We connect client records in HubSpot, GoHighLevel or InOne CRM, calendars and mail in Google Workspace or Microsoft 365, accounting in Xero or Sage, payment collection through PayFast, and customer messaging over WhatsApp Business Cloud API or Twilio.
The systems the business already trusts stay the source of truth. Data connectors read and write to SQL and NoSQL databases, ERPs and custom apps through secure APIs, queues and webhooks. Anything that needs its own home lands in Supabase or PostgreSQL, and everything runs behind Cloudflare. If a tool has an API, an agent from the AI marketplace can usually talk to it. If it does not, we say so before any build starts rather than after the invoice.
Is the AI marketplace POPIA compliant, and who approves what?
The AI marketplace is POPIA-aware from the first design session, because an AI employee touches customer data on its very first shift. Consent is captured explicitly, with the source and the time stamp recorded. Every automated message carries clear opt-out wording, and template usage is logged so an audit can show what was sent and when.
Access is role based, with SSO and SAML available for larger teams, and audit trails record who touched what. Each journey collects only the fields that journey needs, retention controls delete records on time, and data processing agreements are in place before go-live. Data is encrypted in transit and at rest, and webhooks are signed. Risky actions wait for a named human to sign off, so nothing sensitive leaves the business unreviewed. A banned claims list keeps automated wording inside the boundary the company sets, and human edits are always preserved.
How does a company start hiring AI employees?
Starting with the AI marketplace is a conversation, not a contract. Pick one outcome first: response time, booked appointments, quotes returned, or cost to serve. Then choose the single agent from the catalogue that moves that number. That conversation costs nothing and usually takes under an hour.
Next we connect the channel and the CRM, ground the agent in the company's own documents, and get the wording approved before anything sends. A pilot runs two to four weeks on the company's own accounts, with SLA timers, alerts and a weekly changelog so progress is visible rather than claimed. Winning variants are promoted, and the next role is hired once the first one is trusted. Then come the playbooks, the training and the handover, so internal teams run and extend the automations themselves. The company owns everything we build: workflows, prompts and data.
Related capabilities. The same parts, your business.
Keep reading. Pages close to this one.
Tell us which job to cover. We ship the agent that does it.
Send one message describing the job that keeps slipping, whether that is reception, lead follow-up, quoting, booking or back-office data entry. We reply with an honest read on which AI employee fits, and what it will take to put it on shift.