AI and data that make it into everyday work.
- Vendor-neutral
- GDPR and EU AI Act planned in technically and organisationally from the start
- Technical support during implementation
What is on your mind right now?
Choose a question to open the matching service.
- We do not know where AI pays off for us.
- We have AI tools but little benefit.
- AI is used in the team without clear rules.
- We can hardly recognise AI-generated content.
- Our processes take too much time.
- We are looking for the right software.
- Our software is not used as planned.
- We need a new website that is fast, accessible and data-minimising.
- Knowledge is lost when people leave.
- Connections in our data are not visible.
- Our data sits in separate systems.
- We collect data but hardly use it.
Services in three fields
Every service starts with a concrete question from your daily work and ends with a result you can use.
AI
Using AI where it pays off
How you notice it
- There are many ideas but no clear priority.
- Pilot projects never get past the test phase.
- It is unclear whether data and systems are ready for AI.
What we do
- Collect use cases with the business units and rate them by benefit and feasibility
- Check data, systems and processes for their suitability for AI
- Test the most promising case with real data and support the implementation technically
What you get
- A prioritised list of use cases with reasoning
- A use case that has been tested with your data
- A clear decision: continue, adjust or stop
Starting pointAI potential workshop, 1 day
Enquire about this topicAI
Getting more out of AI
How you notice it
- There are licences for AI tools, but they are hardly used.
- Results vary widely from person to person.
- Nobody knows exactly which tasks are suitable.
What we do
- Training by role: management, business units, power users
- Working on the team’s own tasks instead of textbook examples
- Templates, guides and assistants for recurring tasks
What you get
- A team that uses AI safely and purposefully
- A guide with proven use cases
- Confirmation of attendance for everyone trained
Starting pointHands-on AI training, ½ to 1 day
Enquire about this topicAI
Security and AI use
How you notice it
- Company data ends up in private AI accounts.
- There is no rule on which data may go into AI tools.
- It is unclear how to implement the GDPR and the EU AI Act technically and organisationally.
What we do
- Inventory: which AI tools are used for what
- A template for your AI policy, understandable and workable in daily practice; the legal review stays with your legal advisers
- Approved alternatives, on request with local or European models
- Review of AI applications for risks such as prompt injection and data leakage
What you get
- An overview of tools in use and their risks
- An understandable template for your usage policy
- A safeguarded solution that people like to use
Starting pointAI security check, 1 to 2 days
Enquire about this topicAI
Recognising AI-generated content
How you notice it
- Fake images, voices or documents look deceptively real.
- Phishing messages have become more convincing through AI.
- It is unclear how to handle suspicious content.
What we do
- Explain clearly how AI generates text, images, voices and videos
- Practise with real and generated examples: characteristics and typical flaws
- Define checks and reporting channels for suspicious content
What you get
- Staff who check more carefully and spot deception more often
- A checklist for handling suspicious content
- Confirmation of attendance for everyone trained
Starting pointAwareness training, ½ day
Enquire about this topicProcesses and software
Optimising workflows
How you notice it
- Data is copied by hand between systems.
- The same information is maintained in several places.
- Approvals and follow-up questions delay simple tasks.
What we do
- Map processes with the people involved and make time sinks visible
- Eliminate manual re-keying and connect systems via interfaces
- Automate routine tasks, supported by AI where it makes sense
What you get
- A process map with concrete improvements
- Implemented automations with documentation
- Measurement of time savings in a before-and-after comparison
Starting pointProcess check, 2 to 3 days
Enquire about this topicProcesses and software
Selecting and introducing software
How you notice it
- There are several vendors but no clear basis for a decision.
- New software has to work with existing systems.
- Previous rollouts took longer than planned.
What we do
- Clarify requirements with business units and IT
- Compare and assess solutions in a vendor-neutral way
- Plan the rollout and connect existing systems
What you get
- A requirements catalogue and a reasoned recommendation
- A rollout plan with clear steps
- Software that works with your systems
Starting pointRequirements workshop, 1 day
Enquire about this topicProcesses and software
Improving software adoption
How you notice it
- The software is not used as intended.
- Spreadsheets and workarounds grow next to the system.
- Training only works for a short time.
What we do
- Analyse usage and processes in daily work
- Simplify settings and processes in the system
- Training with lasting effect, with contact persons in the team
What you get
- A clear overview of where things get stuck and why
- Simplified processes in the system
- A team that likes using the software
Starting pointUsage analysis, 2 days
Enquire about this topicProcesses and software
Custom website
How you notice it
- The current site is slow, hard to maintain or not accessible.
- Fonts, maps or videos come from third parties without this being clarified.
- Content can only be changed through an agency or developer.
What we do
- Clarify content, audiences, structure and languages
- Develop the site individually: statically pre-rendered, with self-hosted fonts and no third-party embeds without consent
- Accessibility to WCAG 2.2 AA (target: BITV), contact form with encryption and privacy notice, cookie notice only where needed, imprint and privacy page prepared as a structure
What you get
- A website in your languages that is readable even without JavaScript
- Hosting in the EU or on your own server
- A handover with documentation so you can maintain content yourself; your legal advisers review the legal texts
Knowledge and data
Keeping knowledge in the company
How you notice it
- Important knowledge sits in a few heads.
- Key people are about to retire or move on.
- New employees take a long time to find answers.
What we do
- Identify knowledge holders and critical topics
- Capture knowledge in conversations and link it to documents and data
- Build a knowledge assistant that answers with sources
What you get
- A structured knowledge base that can be maintained
- Answers with evidence of where they come from
- Shorter onboarding and fewer follow-up questions
Starting pointKnowledge check, 1 day
Enquire about this topicKnowledge and data
Building knowledge graphs
How you notice it
- Only individual people know how data is related.
- Search returns documents, not answers.
- AI applications need a reliable knowledge foundation.
What we do
- Clarify whether a knowledge graph makes sense for you
- Develop the target picture and a model of terms together
- Connect data and documents and keep the graph traceable and up to date
What you get
- A model of terms and a target picture
- A knowledge graph with a source for every statement
- A foundation for search, analytics and AI
Starting pointKnowledge graph workshop, 1 day
Enquire about this topicKnowledge and data
Connecting data
How you notice it
- ERP, CRM and file shares know nothing of each other.
- Simple questions require searching several systems.
- Numbers contradict each other depending on the source.
What we do
- Create a system map: which data lives where, who uses it
- Agree on shared terms and relationships
- Connect systems via interfaces or a knowledge graph, without vendor lock-in
What you get
- An overview of systems and data flows
- One reliable view of related data
- A foundation for analytics and AI
Starting pointData map, 2 to 3 days
Enquire about this topicKnowledge and data
Using data
How you notice it
- Data is collected but hardly analysed.
- Reports are assembled by hand in spreadsheets.
- Decisions rely more on gut feeling than on numbers.
What we do
- Clarify the questions that really matter to the business
- Check and improve data quality
- Build analyses, dashboards and forecasts, including natural-language queries
What you get
- KPIs that update themselves
- Better data quality as a foundation
- Decisions that can be justified
Starting pointData workshop, 1 day
Enquire about this topic
What sets us apart from traditional consulting
Consulting from practice
Recommendations come from someone who develops software and knows the technical details, and they still stay easy to understand.
Technical know‑how in implementation
We support the implementation with technical know‑how, from connecting systems to handing over to your team.
Traceable
Every recommendation comes with reasoning. Where possible, AI solutions answer with sources.
Vendor-neutral
We recommend what fits you, even if it does not come from us.
Your team becomes independent
Documentation and handover are part of the job. You stay independent of us.

Noah Kirsch
Founder, software developer and consultant · Your contact
I am the founder of Phoenix Knowledge Engineering, a software developer and consultant. My focus is software that connects data, workflows and decisions, from consulting to products of our own such as PubOS.
I bring this practice into consulting. I know the technical details behind AI applications, interfaces and data models and translate them into clear recommendations.
What matters to me is that solutions work in daily practice and are easy to use.
“Technology should make work easier, not more complicated.”Talk to me directly
How we can work together
You can start small and decide how to continue after the first result. We adapt scope and content to your team.
- Initial consultationapprox. 30 minutes · no obligationWe clarify goal, situation and scope and suggest a suitable starting point.Enquire
- Workshop1 day · on site or remoteTogether with your team: assess use cases, processes or data. The result is a prioritised list of next steps.Enquire
- Training or seminar½ to 2 daysUsing AI safely and effectively, basics of data and knowledge graphs, always with cases from your daily work.Enquire
- Implementation supportdepending on scopeTechnical support during implementation, from connecting systems to handing over to your team, with documentation.Enquire
- Talk45 to 90 minutesExpert input for executives and business units, clear and with examples.Enquire
Example topics for talks and seminars
- AI in the company: possibilities and limits
- Avoiding shadow AI: rules that work
- Securing knowledge before it retires
- From scattered data to connected knowledge
- Automating workflows with and without AI
- Recognising AI-generated content: deepfakes, fake texts and phishing
- Using data: from spreadsheet to KPI
- Knowledge graphs: basics and applications
Four steps to a result
Initial consultation
Clarify goal, situation and scope.
Assessment
Analysis of systems, data, processes and existing knowledge.
Recommendation
Clear recommendations with reasoning.
Implementation and handover
We support the implementation technically and prepare your team to carry on by itself.
Mid-sized and large companies
Before you ask
Your question is not listed? Write to us.
How does an initial consultation work?
The initial consultation is without obligation. We clarify goal and scope, then you receive a proposal with a fixed scope.
Do we have to use software from you?
No. Consulting is vendor-neutral. We offer the PubOS software ourselves. If a recommendation concerns our own product, we say so openly. We only recommend our own software if it fits your task.
Is this worthwhile for smaller companies too?
Yes. Small steps with visible benefit matter most there, such as an automated process or training for the team.
Can AI solutions run in Europe or on our own premises?
Yes. Depending on the requirements we recommend European cloud services or local models on your own hardware, so your data stays under your control.
Does advice on the GDPR and the EU AI Act replace legal advice?
No. We advise on technical and organisational implementation. We do not provide legal advice; the legal assessment stays with your legal department, your data protection officer or a law firm.
Do you work on site or remotely?
Both. We like to run workshops on site, much of the rest works well remotely.
Which topic is on your desk?
A short initial consultation clarifies goal, format and scope. Terms are agreed afterwards.

Your contact
Noah Kirsch
Founder, software developer and consultant
Or directly by email: noah.kirsch@phoenix-knowledge.de