APPSS · Application server for AI-native business

You want to become an AI-native organisation, right?

A customer asks where their order is. A supplier confirms a different date. Finance spots an invoice discrepancy. APPSS connects your systems, people and agents around the same business rules and makes the agreed tasks available as services.

Discuss your first step View the references

One idea, your name and email. Then we get to know each other.

Explore everyday business tasks

Your digital team · AI roles

Teamwork starts with shared knowledge.

A team shares goals, knowledge and responsibilities. Our AI roles ask questions, make proposals and challenge each other. Your people bring experience and judgement.

Finn

AI agent · The questioner

“What are we trying to achieve together? What do we still need to know? I would rather ask than finish an unclear task at record speed. You keep the coffee.”

Dex

AI agent · The builder

“You say, ‘Could it work like this?’ I say, ‘Love it!’ Sam says, ‘Let’s check first.’ We keep each other on our toes.”

Noor

AI agent · The organiser

“I keep everyone on track. Very politely, of course. But ‘I’ll get back to you’ does earn a place in my calendar.”

Sam

AI agent · The guardian

“Your business sets my rules of engagement. APPSS enforces what I may access and do. My enthusiasm does not grant extra permission.”

Pip

AI agent · The designer

“Your logo, your colours. Your CRM should not look as though you borrowed it from next door.”

Ari

AI agent · The architect

“Customers, appointments and invoices belong together. I even have a diagram. Surprising, I know.”

Ravi

AI agent · The data doctor

“The customer has the same name here as in accounting. An excellent place to start.”

Tess

AI agent · The challenger

“I ask the question nobody wants: ‘What if this goes wrong?’ After that, I’m usually lovely company again.”

Lou

AI agent · The steady hand

“AI moves fast. You do not have to chase every new thing. I like keeping things together while we keep building.”

AI agent · The guide

“You do not need to become a programmer. Just tell us what you want to remember. Coffee is still welcome.”

Simpler work: People and agents work from shared business knowledge and agreed goals.
Reduce costs: Build on the same APPSS engine, with data storage included. Fewer separate solutions to maintain.
Better communication: Service sees the current order. Purchasing reviews a changed delivery date and Finance sees which invoice needs attention.

Which business task would you like to improve?

Start with work that comes up every day. Connect the systems and people needed to complete the task.

Van Donzel

Check order status

DataInbox looks up the current order status in iONE. A staff member takes over if the available information is insufficient.

Mertens

Release an order

The order from the ERP system and the customer record from the CRM come together. The order can proceed once all required checks have passed.

Xenos

Post and track a transaction

Rabobank transactions from all stores are posted to the relevant general ledger accounts in AFAS. Processing is auditable from transaction to ledger entry.

Browse all business tasks

References

Explore applications

Reference

Riveer

AI agent KAI for tickets, payments and travel information on Ferryon.

“What we particularly value is the flexibility: as a ferry operator, we can manage so much ourselves in the app.”

Riveer · ferry operatorTranslated from Dutch

More about Riveer

Reference

Regiotaxi Noordoost-Brabant

Wmo passes and trip records from multiple operators come together in one complaint file.

Reference

Mertens

Ongoing KYC between ERP and CRM, with people and agents.

More about Mertens

Reference

KOEK

Internal questions and knowledge, with KOEK shaping the experience and APPSS as backend.

More about KOEK

Reference

DailyNexus AI

Your news and knowledge, with explainable AI selection.

Visit DailyNexus AI

Reference

Padelracket.ai

Racket advice from the retailer's own catalogue.

Visit Padelracket.ai

Reference

HealthQIT

Care pathways with healthcare professionals and Panton.

Visit HealthQIT

Trust by design. Your business rules define who can do what.

You define the data and actions that people and agents need for each task. The APPSS server applies those rules. To answer a delivery question, for example, an agent receives the order status and delivery date. Payment details stay out of the conversation.

From configuration to daily work

Configure and test with agents on the left. Choose your interface above the APPSS runtime. Use your template library and swappable AI models for your business context on the right.

Configure with agents

Your business rules, privacy agreements and AI policy become testable configuration.

Agent roles with skills and project context

DexPrepare configuration
TessTest and compare
SamReview permissions and controls

Your chosen AI tools help configure

Codex
Claude Code
Gemini

App skills
CLI
MCP
API

GDPR · AI Act · your own policy

Validate, test and compare. Your authorised team approves changes.

Execute on the runtime

Your AI interface
ChatGPTClaudeGemini
Talk to APPSS, ask questions and create reports.Through configured MCP tools
Webchats and messagesDataInbox
Your frontendWeb, app or portal
Authorised executing agentsFollow-up work within your agreements

APIs and configured MCP tools

APPSS runtime
application server

Your data, business rules and process state remain the basis for every action.

Identity
Permissions
Rules

The runtime checks every action. Human approval where agreed.

Templates and model choice

Your template library

Start with a template from your existing library and configure it for your business process.

Sales
Service triage

Hot-swappable AI models

Choose one or more models for your business context.

Cloud
Private
Local

Change the model connection, keep your business rules. Suitable models and connections are configured and verified for each application.

Your interface, templates and model choice grant no additional permissions. The runtime still checks every business action.

Configuration agents help define and test your agreements. Authorised executing agents do follow-up work within those same agreements.

Where do GDPR and the AI Act fit?

Define purpose, data use, access and retention. Translate relevant privacy agreements into configuration and control points.
Assess the AI use case and its risks. Define the required transparency, human oversight and responsibilities.
Test configured rules, exceptions and permissions. Review changes and retain appropriate evidence.

Applicability, legal assessment and the required technical and organisational measures are defined and verified for each application.

EDPB: privacy by design European Commission: AI Act

How do agents already help test through the CLI?

chat validateChecks a dialogue suite locally without sending messages.
chat runRuns agreed test dialogues on an explicitly selected channel and records replies. This is a live test that can activate agents and tools.
chat compareCompares two test reports as a packet for review. It does not automatically approve content.

These are existing Datainbox CLI capabilities for the chat layer. Agents help review the results. APPSS business rules, the frontend and real effects also need their own project tests; a chat report does not establish those.

Explore the rules of engagement Explore integrations

New AI. Same way of working?

A new connection does not transform an organisation.

You give an agent a task. It uses a connection and returns an answer. But who shares the context, reviews exceptions and looks after the whole process? Without clear answers, your team keeps switching between isolated requests and systems.

Start with one process. Your people and our domain experts make goals, business knowledge and responsibilities explicit. The agent team helps configure and test the application on APPSS. Your people review how it works; together, we decide the next step.

Which situation do you recognise?