Skip to content
ingenit

Development

Custom software development

Software for processes no off-the-shelf product can handle – built from scratch or modernized from a legacy system. With the experience of almost 30 years of project work and the speed of AI-assisted development.

Stock photo: source code on a screen

How we work

Five steps, no black box: You’ll always know where your project stands, what’s next, and what the cost range is.

01

Analysis & assessment

Processes, existing systems and goals – understood, documented, evaluated. AI helps us get to grips with existing code faster.

02

Concept & architecture

A concept you can base decisions on, with a clear technology and cost basis. This is where experience matters most.

03

Development (AI-assisted)

Our computer scientists write code together with AI agents: the agents deliver a first draft and routine work; we steer, write code ourselves and review.

04

Review, testing & approval

Code review, automated tests and your acceptance – for every line, no matter how it was created.

05

Operations & ongoing development

Stable in operation – at your site or in our data center – and ready to grow with your requirements.

What we offer

Six building blocks, one standard: software built to last.

New custom software

Web applications, portals, and interfaces – for planning, billing, operations management, sales, administration, or whatever else you can think of. With AI, we can get to a first working prototype faster; we build your software so that it will still be maintainable even ten years from now.

Modernization & migration

Moving legacy systems to modern platforms step by step: refactoring, new user interfaces, migration of data and operations – without disrupting your day-to-day business.

System assessment & architecture consulting

An honest assessment of what your existing system delivers and what the next steps will cost – including an evaluation of where AI will actually help you. The foundation for every investment decision.

Operations & maintenance

Operations within your infrastructure, with a cloud provider, or at our data center in Dortmund: monitoring, updates, support—and continuous improvement based on ongoing operations.

Customizing standard & open-source solutions

Not everything has to be built from scratch: together with you, we select proven solutions, then customize, extend and integrate them – preferably open source, such as Nextcloud, TYPO3 or Wazuh.

AI in your application – on-premise too

Want your application to have AI features of its own, for example for search, analytics or document processing? We integrate the right model – in our data center, as a provider's frontier model, or locally at your site: if the application is to run entirely in-house, we set up the AI models on your infrastructure and support them in operation.

AI-assisted development

Experience decides what gets built. AI makes it happen faster.

At our company, people and AI work together – guided by experienced developers. We use AI agents where they truly help us move forward, and we don’t use them where sound judgment is required. The result: shorter project timelines and more time for our experienced developers to focus on the issues that determine a project’s success.

How we work with AI agents
  1. 1 Human

    Scope the task

    Define the goal, context, and boundaries – what should the agent deliver, and what should it not?

  2. 2 AI agent

    First draft

    Code draft, architectural ideas, test cases for greater coverage, technical documentation.

  3. 3 Human

    Evaluate & build on it

    Discard, revise, write code yourself – or another round with the agent.

  4. 4 Human

    Review, tests & approval

    Just like any other line of code: code review (including the dual-review process), static analysis, and automated tests.

What AI agents do for us

  • A first draft for new features
  • Ideas for software architecture
  • Test coverage: test cases and test data
  • Technical documentation
  • Recurring patterns, boilerplate code and migration scripts
  • Making sense of code that has grown over the years
  • Suggestions for refactorings

What our people do

  • Understand your domain and ask the right questions
  • Architecture and technology decisions
  • Steer the agents: scope tasks, evaluate results, refine
  • Write code themselves – especially where it gets demanding, both in domain logic and technically
  • Review, test and approve every single line
  • Security, data protection and operations
  • Take responsibility toward you

Noticeably faster

Faster prototypes, shorter analysis phases for legacy systems, more reliable estimates. You benefit directly: earlier results and more predictable budgets.

Skills that last

To judge AI-generated code, you have to be a good programmer yourself. That is why we don't leave everything to the agents: our developers keep writing code themselves – deliberately, so their skills don't erode. Every review relies on that.

Human accountability

Every line of code – whether written by hand or generated by AI – undergoes a code review, static analysis, automated testing, and approval by developers with degrees in computer science.

You decide where the AI runs.

For development as well as for AI features in your application: we adapt to your requirements for data protection, performance and operations.

Default

Local in Dortmund

Our local models run in our own data center. Source code, data models and domain knowledge never leave our building and never end up in training data.

At your site

On-premise

If your AI-enabled application is to run entirely in-house, we set up local models on your infrastructure – for example with vLLM and Open WebUI, optionally with RAG over your documents and tools via MCP servers. Sized to fit and supported by us in operation.

On request

Frontier models

Want to use the most powerful models from OpenAI, Anthropic or Mistral? Then we use them – agreed per project, with a data processing agreement and terms that rule out training on your data.

For procurement & data protection:

Our five binding rules for using AI – short, verifiable, documentable.

Modernization & migration

Your legacy system is running. The only question is: for how long?

Systems that have grown over time often contain decades of domain knowledge. We make sure it isn't lost during modernization – but lives on in a platform that can once again be maintained and further developed.

Stock photo: old computer in the office

Sound familiar?

  • The vendor has discontinued development – or it has become barely affordable.
  • No one in-house really knows the system anymore.
  • New features are waiting on a vendor's schedule.
  • The platform is so old that hardly anyone wants to maintain it – or can.
  • It still runs – but every change is a risk.
  • Maintenance contracts renew automatically; the budget doesn't.
  • “Big bang” migration offers sound like a lottery ticket.
  • The question “What is the system actually worth?” remains unanswered.

Three paths – which one fits depends on your system.

Path 1

Refactoring & stabilization

The system stays, but becomes understood, tested and maintainable again. Low risk, quick impact.

Path 2

Step-by-step redevelopment

Rebuilt module by module, old and new run in parallel, data is migrated gradually. Our most common path – and the gentlest on your day-to-day business.

Path 3

Consolidation & replacement

Several legacy systems are merged into one modern system – including data cleansing and harmonized processes.

Why it's predictable with us

We have been modernizing systems that grew over decades for many years, and we know where the surprises are. Today, AI helps us get to grips with even large existing codebases much faster – locally by default, without your code ever leaving our premises. That's why you don't get a gut-feeling estimate, but an effort range based on solid evidence.

Customizing solutions

Not everything has to be built from scratch.

Sometimes a proven off-the-shelf solution is the better starting point – ideally open source. We select it with you, adapt it to your processes, extend it with whatever is missing and integrate it cleanly into your system landscape. Our focus remains custom development – but we only recommend it when it pays off for you.

Standard where it fits. Custom where it counts.

Select

Market overview, requirements matching, an honest make-or-buy recommendation.

Customize & extend

Configuration, custom modules and plugins, user interfaces built around your workflows – implemented to stay update-safe.

Integrate

Interfaces to ERP, CRM, accounting or single sign-on – data flows instead of being maintained twice.

Operate

At your site or in our data center in Dortmund – including updates and security patches.

What we work with, for example

A selection – and much more.

Collaboration & knowledge

Files, calendars and internal wikis – in-house.

  • Nextcloud
  • Outline

Websites & CMS

Including custom extensions where the standard falls short.

  • TYPO3 + custom extensions
  • WordPress

IT operations & security

Infrastructure documentation and security monitoring.

  • NetBox
  • Wazuh
  • OpenSearch (SIEM)

Local AI platforms

With RAG over your documents and tools via MCP servers.

  • vLLM
  • Bifrost
  • Open WebUI

Why we prefer open source

It fits what we stand for: control over your own software.

  • No per-user license fees
  • No vendor lock-in
  • Source code open to inspection and audit
  • Can be operated in-house or in Dortmund

Let's talk about your project.

30 minutes, free of charge and without obligation: we listen, ask the right questions and tell you openly what we can do for you – and what we can't.