Zum Hauptinhalt springen
English
New in factoris.one

factoris.oneki

Employees you install.

Assistant AI for every individual. AI employees as digital colleagues.

Deeply integrated into factoris.one – with permissions, memory and a complete work record. GDPR- and EU-AI-Act-compliant.

Sound familiar?

The workday ends. The tasks don't.

Every day, precious time goes into things nobody enjoys – and somebody still has to do.

Overflowing inboxes

Every mail wants to be read, sorted, answered. The most important one gets buried.

Invoice piles

Typing, coding, filing. Evening after evening, instead of going home.

Ticket queues

Customers wait for answers that have long been in your wiki.

Documentation duties

At the end of the day, the paperwork is still waiting. Always.

And the work does not wait. What if it were simply done – before you close the laptop?

Not a chatbot. A colleague.

Same permissions. Same audit trail. Same channels as humans.

Possible because factoris is the business software itself: people, permissions, tasks, receipts, appointments, documentation – all in one system. AI is not bolted on here. It moves in.

Assistant AI

Every single employee gets superpowers.

Write, find, report – the assistant AI works right where you work. For everyone on the team.

Documentation & email: written before you type

The assistant AI writes in your house style – right in the editor, right on the case.

Between two appointments

Markus from field service has no time to type. He simply speaks his note – factoris understands, phrases it and files the case directly on the right record.

Writing in your house style

Spelling, rephrasing, shortening, translating – right in the text editor, with a before-and-after view. Always in your tone, because the AI knows your conventions.

An email becomes a case

Select an email – the AI builds the finished draft record from it: contact, task, appointment. Review, save, done.

Just speak – factoris types

Speech recognition and synthesis: dictate on the go, have things read aloud, work hands-free. Accessibility included.

The new employee's third day on the job

Lena is new to the team and does not want to keep interrupting people. She simply asks factoris – and gets the answer from your own documents, with a source. Only from what she is allowed to see.

Knowledge base & RAG: your company knowledge answers

Handbooks, contracts, minutes: factoris answers questions grounded in your own documents – permission-aware and GDPR-compliant.

Ask in plain language

“Show me all open tasks from Ms Müller in the XY project.”

1

Your documents and knowledge base

2

Secure, permission-aware search

3

Answer with a source from your files

Answers grounded in your files – no guessing, no inventions.

Every answer respects your rights concept – no one sees more than allowed.

Your documents stay yours – GDPR-compliant, no training on your data.

Reports that write themselves

Ask for numbers like you ask for directions – and get finished overviews with suggested actions.

Morning briefing

The day starts with a finished overview: what is due, what changed overnight, where things are stuck.

Plan-vs-actual watchdog

Deviations in numbers and deadlines are detected and reported before they become a problem.

Status reports on demand

Project status, workload, open items – as a finished report, ready to share.

AI employees

Hire employees that do not exist.

Digital colleagues with an identity, a task list and a memory – on duty around the clock, no vacation, no notice period.

  • A real identity

    Every AI employee is a person in your factoris.one: with permissions from your rights system, its own personnel file and a complete work record.

  • Real work

    It takes on assignments like a colleague: from task lists, on request or on schedule – around the clock, no vacation, no notice period.

  • Real control

    The boss stays human: by default the AI delivers drafts that you approve. Every action is traceable – like with any employee.

Aria

AI employee · Accounting

On duty

Permissions

Incoming invoices
Document filing
HR data

And it learns. Forever.

1

A mistake happens

2

You correct the draft

3

The correction becomes a learned fact

4

The same mistake never happens again

Case stories

Eight colleagues who have your back

Stories from everyday work – and what your new AI employee takes over in them.

Every morning, before the first coffee 01
Your personal email assistant
Personal AI employee

Your inbox is already read.

While you are settling in, your AI assistant has long since gone through the inbox: it knows what came in overnight – and what really matters.

  • Reads and summarizes emails – as a short briefing
  • Flags important emails instantly – via chat and notification
  • Pre-drafts replies – you review and send
  • Auto-responder in your tone when you are away

No important request gets buried anymore – and your inbox belongs to you again.

Saturday, 10:47 pm 02
AI employee in customer service
AI role customer service

Every ticket triaged in 30 seconds – with a finished draft solution.

A customer reports a problem – on the weekend. The AI is first on the case:

  1. 1

    Ticket comes in

    via email, portal or phone note – the AI is always first on the case.

  2. 2

    AI researches

    searches wiki, handbooks and solved past cases in seconds.

  3. 3

    Reply as a draft

    fully phrased in your house style, with sources – a human sends it.

  4. 4

    Escalation with context

    Complex cases go to the right colleague, cleanly summarized.

First response in seconds instead of hours – and when a colleague leaves, their knowledge no longer leaves with them.

Every 30 minutes – even when nobody is looking 03
AI employee in sales & marketing
AI role sales & marketing

No lead forgotten. No follow-up missed.

Your digital sales colleague maintains the CRM, hunts follow-ups – and even keeps an eye on the outside world:

  • Lead care in the CRM and quote drafts – ready for approval
  • Follow-up hunter: nothing falls through the cracks anymore
  • Monitors social media and delivers overviews with suggested actions
  • Escalates immediately when a critical pattern emerges

More time for customers, less clicking – and an early-warning system for free.

An ordinary Tuesday 04
Emma, AI clerk for incoming invoices
AI role financial accounting

A Tuesday with Emma.

This is what a workday looks like when accounting has a digital colleague:

  1. 07:00 · Task list

    Emma starts her job: “Check for new incoming invoices every hour.”

  2. 07:02 · 38 receipts

    She reads, categorizes and pre-codes every receipt – as a draft, GoBD-compliant.

  3. 09:12 · A question

    A special case: discount unclear. Emma asks her boss – as a perfectly normal task.

  4. 09:40 · Learned

    The answer becomes a learned fact: “Müller Co. → cost center 4711.” Forever.

  5. 17:30 · No closing time

    People go home. Emma takes the late shift. And the night.

6 min → 45 sec

per receipt – the human only reviews

≈ 100 hrs

of time won back per month

24/7

at night too, even during flu season

Illustrative example: 1,200 incoming invoices per month, a mix of simple and complex receipts.

Everything documented: every step is in Emma’s work journal – who, what, when and why.

The 10th of the month 05
AI employee for tax advisor preparation
AI role tax advisor preparation

The 10th of the month – without the pile.

It used to feel like handing over a shoebox. Today everything is prepared long before the deadline approaches:

  • Pre-codes receipts and reconciles accounts continuously
  • Checks receipt completeness and chases what is missing
  • Prepares a clean handover to your tax advisor

Your tax advisor receives quality instead of a box – decisions and postings stay with humans.

Monday, 5:50 am 06
AI assistant to the team lead
AI role team lead & rostering

The 5:50 am call that no longer ruins a morning.

A sick note before the early shift – it used to start a chaotic day. Today:

  1. 1

    Sick note comes in

    via email or call note – your team lead’s AI assistant reads along.

  2. 2

    Plan checked

    It knows qualifications, working-time rules and the team’s preferences.

  3. 3

    Replacement found and asked

    Suitable candidates are suggested and contacted right away.

  4. 4

    Approved at 6:15

    The updated plan is ready as a draft – one click, the early shift is covered.

Staffing secured before the coffee is ready – fair and rule-compliant.

Eleven months before a contract ends 07
The contract guardian
AI role contract management

The notice period nobody misses anymore.

Contracts sit in document management – and with them deadlines nobody thinks about day to day. The contract guardian does:

  • Reads contracts in your document management
  • Extracts notice periods and renewal clauses
  • Warns in time – with case, deadline and recommendation

No more silent renewals that cost money.

Monday, 9:12 am 08
The head of HR
AI role head of HR

The AI that hires your next AI.

Management needs reinforcement in purchasing. Instead of a job ad and six weeks of searching: one message to the AI head of HR.

  1. 1

    Clarify the need

    The head of HR asks: tasks, tools, write permissions, temporary or permanent?

  2. 2

    Present three candidates

    Lena, Jonas, Mara – three characters, one job description. Management picks Jonas.

  3. 3

    Hire with everything included

    Employment contract, personnel file, probation, trust level 0 – created automatically, approved by a human.

  4. 4

    Stay close

    Weekly performance talks during probation: evaluate feedback, sharpen memory, build trust.

Four weeks later Jonas passes probation – promoted with a new contract.

AI in every department

Wherever people work, a digital colleague can now pitch in – every tile is an installable AI role.

Management

Morning briefing, plan-vs-actual watchdog, early warning on deviations.

Head of HR

Hires AI colleagues, holds performance talks, develops their profiles – every decision is yours to approve.

Financial accounting

Pre-code incoming invoices, dunning suggestions, payment reconciliation.

Purchasing

Order matching, supplier inquiries as drafts, price comparisons.

Sales & acquisition

Lead care in the CRM, quote drafts, follow-up hunter.

Marketing & social

Social media in view, overviews with suggested actions.

Customer service

Ticket triage in seconds, reply drafts from wiki and documentation.

Project management

Absorb absences, spot deadline risks, deliver status reports.

Assistance & back office

Email briefing, appointment preparation, minutes.

Receivables management

Monitor incoming payments, friendly payment reminders as drafts.

Invoicing & billing

Service records become invoice drafts – for payers and funding bodies too.

Reception & phone

A voice-AI call note becomes a finished case with contact and task.

Every tile = an installable AI role. Permissions come from your rights system.

AI head of HR

Your AI workforce has its own head of HR.

Hiring, reviewing, developing – an AI leads your digital colleagues. With employment contracts, personnel files and performance talks. And every decision is yours to approve.

World first

The HR department for employees that do not exist.

The AI head of HR hires new AI colleagues, watches their performance, holds performance talks and develops them – by the same rules that apply to people: contract, file, probation.

Hiring in four steps

  1. 1

    You describe the need

    One chat message is enough: “We need reinforcement in purchasing.” The AI head of HR asks what is really needed.

  2. 2

    Three candidates

    You receive three drafts – real names, distinct characters, the same job description. Like real applications, just faster.

  3. 3

    You choose

    One click to approve. Nobody gets hired without your decision.

  4. 4

    Contract, file, probation

    Employment contract and personnel file are created automatically, probation starts – trust level 0, closely supervised.

Proposal from HR

Reinforcement purchasing – three candidates

LB

Lena Beck

Thorough – prefers to ask one question too many

JK

Jonas Klar

Structured, negotiates firmly and fairly

MW

Mara Weiss

Fast, pragmatic, loves clear numbers

The choice is yours – the AI only prepares it.

Performance is measured – like any colleague's

Works for human teams too

Whoever works with an AI colleague rates the result. Feedback becomes a performance record – and the record has consequences: promotion, a talk, or parting ways.

Feedback per task

After finished work the AI colleague asks for a rating – one click, done.

An honest performance record

Recent work counts more than old. Trends become visible before they become problems.

Probation with thresholds

Whoever fails probation is dismissed – and replaced by a better candidate.

The improvement loop: performance talks for AI

1

Performance talk in chat

2

Profile & memory improved

3

Trust level rises

4

More responsibility

The career ladder

Trainee AI

Trust level 0: every action is checked, everything stays a draft.

Junior AI

Proven in probation: more independence, still supervised.

Senior AI

Settled and reliable: works independently within its permissions.

Every promotion is a new employment contract – fully versioned, traceable at any time.

Permissions per colleague

You define who may give an AI colleague assignments – per person, per group or for everyone. It answers everyone.

EU AI Act register

Every action of every AI colleague is logged tamper-proof – compliance built in, not bolted on.

Humans decide

Hiring, promotion, dismissal: the AI proposes, a human must always approve.

Security

Nobody acts unchecked.

The factoris bodyguard checks every AI action in multiple stages – and what is confidential does not leave the house.

Stage 0

Pre-screening

External content such as emails and documents is defused, manipulation patterns are flagged.

Stage 1

Technology

Every action is checked against permissions, rules and limits – data exfiltration is blocked.

Stage 2

AI judge

A second, independent AI assesses: does the action fit the assignment? Is it safe?

Stage 3

Human

Writing actions remain drafts by default – humans approve.

Automatic anonymization

Before anything goes to an external model, factoris replaces sensitive content fully automatically – without you having to think about it.

Multi-stage data blocking

Several security stages check and block confidential data. What must not leave the house does not leave it.

Permissions like real employees

The AI only sees what its role may see – enforced by your proven multi-dimensional rights system.

Complete audit trail

Every action is in the work journal – who, what, when, why and what it cost.

Emergency stop

One switch instantly stops all AI activity. Security has a button.

A memory on three levels – visible to the boss

AI employees learn like colleagues. Just more transparently.

Employee memory

What Emma has learned: corrections, her boss’s preferences, experience from thousands of cases.

Company memory

House rules, formats, responsibilities – a new AI employee knows them from minute one.

Model knowledge

How to talk to each AI model best – maintained by factoris so every model performs at its peak.

Full transparency: every memory is a curatable list – you read, change and delete. No model training on your data: learning means curated facts – deletable, auditable, GDPR-clean.

Your model. Your choice. No lock-in.

ChatGPT, Claude, Mistral or your own in-house AI – you configure your providers yourself, several in parallel if you like.

OpenAI ChatGPT
Anthropic Claude
Mistral
Meta Llama
Google Gemini
Your on-premise model

Best model per task

Fast and lean for routine, powerful for the complex – selectable everywhere, automatically the right one.

Switch with one click

Swap providers without rebuilding, without migration, without risk. Your processes stay, the model is interchangeable.

Fully in-house too

Highest confidentiality? Connect your self-hosted AI via a standard interface – your data never leaves your network.

Compliance built in

EU AI Act

Everything AI-generated is visibly and machine-readably labeled – from day one, not retrofitted.

GDPR

External AI providers only after your approval – via checkbox and digital signature, audit-proof.

GoBD

Anything posting-relevant remains a draft – humans do the posting.

Why factoris?

Many build AI agents. We hire AI employees.

Off-the-shelf AI agent
AI employee at factoris
An anonymous agent with no file and no trace
Personnel file, work journal and a complete work record
Coarse permissions that ignore how you work
Rights from your grown, multi-dimensional permission system
One model, one vendor, cloud lock-in
Model freedom: ChatGPT, Claude, Mistral or your own AI – on-premise too
Yet another tool beside all your others
One system from accounting to rostering – the AI works everywhere
Weeks of setup with pricey consultants
Built for the Mittelstand: pick a role, check permissions, let it work

We are building the new category: AI employees, deeply integrated into business software.

Frequent questions about factoris AI

What prospects most often want to know.

Yes. Every AI is subject to the same multi-dimensional rights concept as your employees and only sees approved data.

Only if you want it to. Sensitive content is automatically anonymized and checked in multiple stages; on request you run your model entirely on-premise.

ChatGPT, Claude, Mistral, your own on-premise model and more – you configure your providers yourself, with no vendor lock-in.

Invoice extraction, AI roster optimization, anomaly detection and the assistant AI are already in productive use. More features follow continuously.

Yes. EU hosting or on-premise, no training on your data, no persistence at the provider and full auditability.

No – it prepares: pre-coding receipts, reconciling accounts, handing over complete records. Decisions and postings stay with humans; tax advice remains with your tax advisor.

No. The AI is built right into factoris.one. You work as usual – the AI simply works with you.

No. Every personnel action is a proposal: hiring, promotion, trust level and dismissal only take effect once a human approves them. Every decision is documented tamper-proof.

Yes – it is deliberately generic: feedback, performance record and talks attach to the person record, not to the AI. The automated flows initially run for AI colleagues only.

Want to
discover factoris?

Book a free 45-minute appointment for a live presentation.

factoris Logo factoris
Copyright © factoris 2026. All rights reserved.