# What AIpril is

> The platform where people and AI employees run internal operations together: applications, processes, shared data and executors that work inside them

Source: https://aipril.dev/docs

AIpril is the platform where people and AI employees run internal operations together. AIpril's AI
turns a business request into a working system: professional applications, the processes that move
work through them, and the shared data model underneath. It then creates the AI employees that
monitor that work, prepare decisions, execute routine tasks and escalate the exceptions.

AI employees are not chatbots layered on top of enterprise software. They work inside the
application and the process, each with a defined role, a scope of authority, a set of data they may
reach, skills, limits and an audit trail. Inside those boundaries they act on their own. Outside
them they hand the work back to a person.

Everything runs on one foundation: one employee directory, one organisational structure, one data
model, one set of access rules. An application, an approval route, a shared workbook, a report and
an AI employee all read the same data and follow the same rules.

## What it replaces

AIpril replaces fragmented SaaS, overpaying for people to do routine work, and spending time and money on external developers, with one governed system: shaped to your process, run with AI employees, improved by your internal team, and enriched by the knowledge they build over time.

### • Software that does not fit

Companies combine several SaaS platforms to run one process. Each tool covers part of the work and
none reflects the full operating model. Teams pay for capabilities they do not need, customise the
stack around its gaps, or change the process to fit what the software supports.

AIpril creates one system around the process as it actually runs: its roles, data, rules,
decisions and workflows.

### • Overpay for people doing routine work

Employees spend their day moving information between systems, checking records, preparing updates,
chasing requests and compiling reports. The work is necessary. Most of it does not require human
judgment.

AI employees take that layer over. They watch for signals, collect and validate information,
execute repeatable tasks, prepare decisions and escalate what needs a person, so the same operation runs with a smaller team, at lower cost, with faster throughput.

### • Process knowledge that does not compound

Your internal experts know how the operation should work. Every improvement turns into a brief for
consultants or developers, so the knowledge is translated and handed over instead of being kept and
built on.

In AIpril the knowledge becomes part of the application: its data model, rules, workflows, roles and
the instructions its AI employees follow. Each change adds to an internal asset instead of starting
another delivery cycle.

## What you get

### • [AI employees](https://aipril.dev/docs/ai-organisation.md)

Operational executors with a role, a manager or process owner, a budget, skills and a defined
authority. They monitor, collect and check information, prepare decisions, do repeatable work and
escalate exceptions. Their actions run through the same permissions, approvals and audit trail as
the actions of people.

### • [Applications](https://aipril.dev/docs/apps.md)

Purchase registers, sales pipelines, service desks, onboarding trackers, project workspaces,
internal portals. AIpril's AI assembles them from governed building blocks rather than writing
unrestricted production code, and each one is built around the process, data, roles and controls it
needs. Changing it later does not require a new project.

### • [Processes](https://aipril.dev/docs/processes.md)

Forms, routes, statuses, deadlines, rules, approvals, exceptions. Every request has a current
owner, so you can see what is happening, who holds it, what comes next and where work has stopped.

### • [Scheduling](https://aipril.dev/docs/scheduling.md)

A project plan recalculated under dependencies and people's capacity, and shift rosters built to
cover demand, by a solver rather than by dragging bars. You see the new plan before it is written,
and you can undo it.

### • [Data](https://aipril.dev/docs/engine.md)

One shared definition of business entities and metrics. Reports, dashboards, applications and Excel
exports draw on the same source, so two teams stop arriving at the same meeting with different
numbers.

### • [Ledger](https://aipril.dev/docs/ledger.md)

Balances in money, points or stock, on double entry, computed once for the whole platform instead
of separately inside every application.

### • [Analytics](https://aipril.dev/docs/analytics.md)

A full BI stack inside the platform, not beside it: exploration, dashboards, cross-filtering and the
chart types an analyst expects, on the same definitions the applications run on, standing next to
the work instead of in a separate tool nobody opens.

### • [Sites and portals](https://aipril.dev/docs/cms.md)

Every screen - the employee portal, an application's own pages, a dashboard, a public page - built
from one widget catalogue on one design system. So your internal screens and your public site
finally look like the same company.

### • [Spreadsheets](https://aipril.dev/docs/spreadsheets.md)

The grid people actually work in, on the platform rather than in a file: collaborative, connected to
real data and to processes, with the formula language they already know. A row can start an
approval; a cell can hold a figure the rest of the system reads.

### • [Master data](https://aipril.dev/docs/master-data.md)

Products, counterparties, warehouses, categories - one list each, of whatever structure you need,
used everywhere instead of retyped per application. Cascading choices in forms, and catalogues that
can be filled from an external service instead of by hand.

### • [Collaboration](https://aipril.dev/docs/collab.md)

The conversation attached to the work: comments on a record, a dataset, a cell range or a project,
mentions that reach people in Slack, Telegram or email, reactions, and a history of what changed.
Your AI employees write in the same threads.

## What holds it together

**One employee directory.** A person is hired, moves department or leaves once. Applications,
routes, reports and access follow from that. There is no second list to update and forget, and no
access left live in the two systems nobody remembered.

**Access lives in the data.** Which rows a person sees and which columns stay hidden is part of the
data model. The rule holds in a table, in a chart, in an Excel export and in an answer from an AI
employee. Security reviews it once, not once per application.

**An application is described, not written.** Its composition sits beside it in a form a person can
read, so it can be reviewed, argued with, compared against last month's version and changed by the
company rather than by whoever assembled it first.

**Everything leaves a trail.** Who changed what, on what grounds, and which version was live at the
time. What an AI employee did lands in the same log as what a person did.

## Where to go from here

- [What you can do with it](https://aipril.dev/docs/overview/what-you-can-do-with-it.md) - the systems a business can
  stand up here, from a CRM to an AI employee that does a real job.
- [How it works](https://aipril.dev/docs/overview/how-it-works.md) - the path from a request to a running application,
  and what happens on each step.
- [Architecture](https://aipril.dev/docs/overview/architecture.md) - the seven layers of the stack, from the AI employees
  down to the infrastructure. The page to forward to IT.
- [Getting started](https://aipril.dev/docs/overview/getting-started.md) - the first hour, in six steps.
- [All capabilities](https://aipril.dev/docs/overview/capabilities.md) - the full index of what is documented.

And the comparisons, if you are placing this against something you already run:
[ERP and SaaS](https://aipril.dev/docs/overview/vs-erp.md) ·
[low-code and BPM](https://aipril.dev/docs/overview/vs-lowcode.md) ·
[vibe coding](https://aipril.dev/docs/overview/vs-vibecode.md) ·
[BI](https://aipril.dev/docs/overview/vs-bi.md).