Available now EzeMind product · proprietary IP

EzeCore: the governed platform for business-specific AI agents

Your workflows. Your AI agents. Your team in control.

EzeCore is EzeMind AI's governed platform for building a department's operating system: one core with secure sign-in and permissions, a dedicated database with security enforced on every record, a permanent audit trail, engines for records, tasks, digital forms, documents, notifications and reports, and an AI agent framework in which agents propose, named people approve and every request is metered. It is licensed monthly, hosted by EzeMind in South Africa or on your own server, and live today on its first departmental deployment.

EzeCore: the governed platform for business-specific AI agents, shown on laptop and phone with department icons
9
Platform layers in every licence
4
Agent kinds: monitor, draft, check, chase
7
Delivery steps from discovery to live
0
Record changes an agent can make without a named approver
The base platform licence

What is in every EzeCore licence

Nine layers, built once by EzeMind and shared by every licensee. Your department is configuration on top of them.

#LayerWhat it doesWhy it matters
1Secure sign-in and permissionsNamed identities for people and agents; roles decide who may see, create, approve or export, down to department and record type.An agent configured for Finance cannot wander into HR files.
2Dedicated database, security on every recordEach business runs on its own database; access rules are enforced at record level inside the database, not only in the screens.A query from an agent, a report or an integration is filtered exactly like a user's screen.
3Permanent audit trailAn append-only log of who or which agent did what, to which record, when, and on whose approval. Users cannot edit or trim it.What an auditor, a regulator or your operations manager reads.
4Document store with secure linksDocuments live on the records they belong to; sharing uses expiring, permission-checked links, including no-login uploads for suppliers.No more attachments in five inboxes.
5Task and calendar engineRecurring and one-off tasks, deadlines, reminders and escalations as first-class objects; one calendar of what is owed to whom.The engine behind "monitor renewals" and "chase outstanding documents".
6Digital-forms enginePaper and spreadsheet forms become structured forms with validation, signatures and photo capture, on desktop or mobile.Agents pre-fill from records and route for approval.
7Records engineRegisters for what the department manages: contracts, suppliers, assets, claims, cases, employees, sites; each with fields, statuses and history.Everything else hangs off these records rather than floating free.
8AI assistant and agent framework with budgetsThe assistant answers from your own records; agents run configured workflows (monitor, draft, check, chase) and propose actions into a suggestions lane. Every agent has daily and monthly AI budgets and a hard cap.Agents propose; a named person approves; the run is logged.
9Export and reporting layerDepartment dashboards, scheduled reports, exports to Excel, CSV, accounting and BI tools; full export of records, documents and the audit trail on any day.Compliance reporting becomes a workflow that assembles, drafts and queues for sign-off.
Agent anatomy

Agents propose. People decide.

Every EzeCore agent works inside the harness. It signs in with its own identity, reads only what its role allows, and writes its proposals into a suggestions lane. A named person accepts, edits or rejects each one; the approved action runs under that person's name through the same engines a human would use; and the whole run is written to the permanent audit trail.

Every request to an AI provider is metered in a usage ledger and capped by a daily and monthly budget. When the cap is reached the agent pauses and the owner is told.

"Agents propose, people decide. That single rule is what makes an AI agent safe enough to put on a department's real records."
Johan van Niekerk, Founder & CEO, EzeMind AI

Monitor agents

Watches registers and dates: contracts nearing expiry, certificates due, documents outstanding.

Draft agents

Prepares the renewal note, the onboarding pack, the report, from the record and the approved template.

Check agents

Validates what arrived: is the invoice complete, does the claim match the policy, are the fields consistent.

Chase agents

Reminds, escalates and follows up on a schedule until a person closes the loop.

Any department. Real results.

Which departments run on EzeCore

Any department with a register, dates that fall due, people who must be chased, decisions that need two signatures and documents that must come from the record.

DepartmentBest forKey advantage
OperationsRenewals, inspections, incident follow-upsNothing due is forgotten; every action is evidenced
FinanceApprovals, reconciliations, claimsTwo-person rules and an audit trail by default
HROnboarding, reviews, leave, certificationsChecklists that survive staff turnover
SalesQuote follow-ups, pipeline hygiene, handoversChasers that never go quiet
LegalContract registers, deadlines, compliance packsDocuments generated from the record, filed before download
Supply chainSupplier onboarding, document checks, delivery trackingSuppliers upload without a login; you see one file
FacilitiesAssets, maintenance schedules, certificatesPhoto-capture forms on a phone, at the site
Property & leasingLease register, renewals, tenant documentsLive today on EzeCore's first departmental deployment
Delivery

Seven steps from discovery to live

Weeks, not years. The core exists and is tested; your department is configuration.

  1. 01

    Discovery call

    Free, 30 minutes. We identify the first workflow and the records and documents it touches.

  2. 02

    Analysis of your file

    We work from your real registers, spreadsheets and documents, not a generic template.

  3. 03

    Decision Register

    Every configuration choice is written down and agreed before the build starts.

  4. 04

    Configure on the core

    Registers, forms, tasks, documents and agents are set up on the harness. No new software project.

  5. 05

    Live demonstration

    You see it running on your own data before any fee is agreed.

  6. 06

    First unit live

    One branch, centre or team goes live completely, with training.

  7. 07

    Bedding in, then the next unit

    Four weeks of bedding in, then the next department or unit follows as configuration.

Pricing model

How a departmental OS on EzeCore is priced

Four lines, quoted after a free discovery call. The licence fee is the same whether EzeMind hosts the platform in South Africa or you run it on your own server.

  • Monthly platform licence. The EzeCore core, maintained and updated. Same fee whether hosted by EzeMind or on your own server.
  • Departmental workstreams. Once-off, priced on value and paid on acceptance, per department configured.
  • Platform Partner service. Monthly: maintenance, security updates, support and a metered AI budget your directors control.
  • Further work. At a published hourly rate, approved in writing before it starts.

Development elsewhere on this site is fixed-price. EzeCore is the exception by design: a licence, because maintenance, security updates and releases stay in the fee.

Where it fits

One-off agent, WhatsApp CRM or departmental OS?

OptionBest forTimelinePrice
One-off AI agentAutomates one repetitive workflow inside one tool1–2 weeksFrom R12,500 once-off
WhatsApp CRMThe channel and database for a customer-facing team4–8 weeksFrom R28,000 once-off
Departmental OS on EzeCoreA whole department's registers, deadlines, approvals, documents and agents on one governed coreWeeks (configuration)Monthly licence, quoted after discovery

Your ERP or accounting system stays the financial system of record. EzeCore is the operating layer above the spreadsheets and below the ledger. An integration to the ERP is promised only after a scoped connectivity assessment proves it.

Licence and data

What you own, where it lives, what we never do

Licensed, not sold

EzeCore is proprietary architecture built and owned by EzeMind (Pty) Ltd and licensed monthly. Your data is yours outright, always; full export of records, documents and the audit trail is available from inside the platform on any day.

Your own database, in South Africa or on your server

A dedicated database never shared with another client, hosted by EzeMind in South Africa or on a server you own and control. Encrypted in transit and at rest, including backups, with nightly encrypted backups and an off-site copy.

Never trained on, always metered

AI requests run under commercial provider terms that exclude training on client data. Every request is recorded in a usage ledger and capped by a budget your directors control; the Confidentiality and Data Processing Agreement signed with the proposal makes it contractual. POPIA is designed in.

EzeCore FAQ

Questions business owners ask about EzeCore

What is EzeCore?

EzeCore is EzeMind AI's governed platform for building a department's operating system: sign-in and permissions, a dedicated database with security enforced on every record, a permanent audit trail, engines for records, tasks, digital forms, documents, notifications and reports, and an AI agent framework in which agents propose, named people approve and every request is metered.

What is a harness, and why does an AI agent need one?

A harness is everything an agent needs around it before it can be trusted with real work: an identity, permissions, a governed data boundary, an approval path, an evidence trail and a budget. Without it, an agent that can read every record and act on any of them is a liability. EzeCore provides all six, enforced inside the database rather than on the screen.

Can an EzeCore agent change a record on its own?

No. An agent proposes an action into a suggestions lane; a named person accepts, edits or rejects it; the approved action executes under the approver's name; and the run, its inputs and the decision are written to the permanent audit trail. Uploaded documents are information only and can never trigger a change.

Which departments can run on EzeCore?

Any department with a register, dates that fall due, people who must be chased, decisions that need two signatures and documents that must come from the record: property and leasing (live today), sales and lead management, logistics and freight, after-sales and warranty, finance and administration, compliance and HR, care operations, and accountability and task management.

How is a departmental OS on EzeCore priced?

Four lines, quoted after discovery: a monthly EzeCore platform licence (the same fee whether hosted by EzeMind or on your own server), once-off departmental workstreams priced on value and paid on acceptance, a monthly Platform Partner service that includes maintenance, security updates, support and a metered AI budget, and further work at a published hourly rate approved in writing before it starts.

Is EzeCore sold outright?

No. EzeCore is licensed monthly so that maintenance, security updates and releases stay in the fee. What you own outright is your data, always, and the right to use your workstreams for as long as the licence runs; full export is available from inside the platform on any day.

Where does our data live?

In a dedicated database that is never shared with another client, hosted by EzeMind in South Africa or on a server you own and control. Data is encrypted in transit and at rest, including backups, and every read and write by anyone, including EzeMind staff, is recorded in the audit trail.

How long does it take to go live?

Weeks, not years. The core exists and is tested; your department is configuration and workstreams. EzeMind analyses your own file, issues a Decision Register, configures on the core, demonstrates it live, then takes the first branch, centre or team live completely, with four weeks of bedding in before the next unit follows.

Is our data used to train AI models?

Never. AI requests run under commercial provider terms that exclude training on client data, every request is metered in a usage ledger, spend is capped by a budget your directors control, and the Confidentiality and Data Processing Agreement signed with the proposal makes it contractual.

What is the difference between EzeCore and an AI agent built as a one-off?

A one-off agent automates one workflow and lives in one tool. A departmental OS on EzeCore gives a whole department its registers, deadlines, approvals, documents and agents on one governed core, under one sign-in and one audit trail, and adds the next department as configuration rather than a new build.

See EzeCore on your own department's file

A free 30-minute discovery call, a written proposal within 2 working days, and a live demonstration on your data before any fee is agreed.

Read more: EzeCore explained, layer by layer · All platforms & apps