AI References – Chatbots & Knowledge Bases in Action
Our AI references cover chatbot, knowledge base and AI automation projects in production – for knowledge preservation, process automation, and employee relief.
AI Projects from Germany: Practical Experience Over Theory
Artificial intelligence is no longer an abstract future topic – it is a production-ready technology that creates measurable value in an increasing number of companies. These AI references show chatbots and knowledge bases in action at real organizations. At Groenewold IT Solutions, we have successfully implemented numerous AI projects in recent years: from intelligent chatbots and RAG-based knowledge bases to specialized machine learning models for industrial applications. On this page, we showcase selected projects that exemplify our approach.
What distinguishes our AI projects is the consistent focus on business value. We do not implement AI because it is technically possible, but because a concrete problem needs to be solved. The AI knowledge base for a machinery manufacturer emerged from the need to preserve the knowledge of long-standing employees before their retirement. The AI cooking assistant Chop-E was developed to offer users a personalized cooking experience beyond simple recipe searches. And the Lullio baby monitor app shows how AI-based sound recognition can create real safety value for families.
Technologically, we rely on a combination of powerful language models (OpenAI GPT-4, Claude, open-source alternatives like Llama and Mistral), modern Retrieval-Augmented Generation architectures (RAG), and proven vector databases like Pinecone, Weaviate, or pgvector. Data privacy is particularly important to us: all our AI solutions are GDPR-compliant. For particularly sensitive data, we offer on-premise solutions that run entirely in your own infrastructure – without external API calls and without data leakage.
Our experience shows that the biggest success factor in AI projects is not the technology itself, but the quality of the data and the clarity of the objectives. That is why we start every project with a structured discovery workshop where we jointly define what problem needs to be solved, what data is available, and how success will be measured. Only then do we select the appropriate technology and develop a proof of concept. This iterative approach minimizes risks and ensures that the AI solution actually delivers the desired benefit.
Our AI Expertise
AI That Solves Problems – Not Just Impresses
We do not develop AI for the sake of AI. Every project starts with a concrete challenge: preserving knowledge, automating processes, supporting decisions. Technology is a means to an end – the value for your company is the focus.
Chatbots & Assistants
Intelligent dialogue systems for customer service, internal knowledge, and process automation
Knowledge Bases
RAG systems that make company knowledge searchable and usable
Machine Learning
Prediction models, classification, and pattern recognition for business processes
Language Processing
NLP for text analysis, summaries, and automatic categorization
Process Automation
AI-powered workflows for recurring tasks and decisions
Privacy-Compliant AI
GDPR-compliant solutions, also with local models without cloud connection
Selected AI Projects
AI Knowledge Base for Machinery Manufacturer
Development of an AI-powered knowledge base for a mid-sized machinery manufacturer. The system captures and structures the expert knowledge of long-standing employees and makes it accessible to all staff via an intelligent chatbot – even after their departure.
Highlights:
Technologies:
Chop-E AI Cooking Assistant App
Development of an innovative AI cooking assistant called Chop-E that helps users discover new recipes and improve their cooking skills. The app offers various search modes such as search by ingredients, by dish, or random dish. With a friendly user interface and personalized recommendations, Chop-E turns cooking into an interactive and educational experience.
Highlights:
Technologies:
From idea to go-live – with a clear roadmap
We structure backlog, milestones and stakeholder communication so your AI project stays predictable: proof of concept, integration with existing systems and handover to operations – without empty promises.
Thorsten Frieling – Project management
Discuss project status
Lullio Baby Monitor App with AI Sound Recognition
The Lullio Baby Monitor App transforms smartphones into a secure baby monitoring solution with HD live video and audio transmission, night vision mode, and two-way communication. With AI-powered sound recognition, the app automatically distinguishes between different baby sounds (crying, laughing, sleeping) and alerts parents intelligently.
Highlights:
Technologies:
AI Phone Assistant for Property Management
Illustrative scenario: A GDPR-compliant AI phone assistant handles routine calls for a property management company – damage reports, meter readings, scheduling, and service-charge questions – around the clock in natural language, routing urgent cases straight to the emergency service.
Highlights:
Technologies:
AI Vision for Quality Control
Illustrative scenario: AI vision on the packaging line of an East Frisian food manufacturer detects product and packaging defects in real time right at the belt – via edge cameras, with no cloud dependency and without slowing the production cycle.
Highlights:
Technologies:
Your AI project – our next reference?
From RAG knowledge bases to AI voice agents: we align use case, data and GDPR-ready architecture – then deliver MVP and scale pragmatically (Made in Germany, Leer/Ostfriesland).
Björn Groenewold – Managing Director
Discuss your AI projectYour AI Project
AI for Your Company – Where to Start?
Whether chatbot, knowledge base, or process automation – we advise you on which AI solution brings the greatest value for your company. Free and without obligation.
Knowledge & Answers
Frequently Asked Questions About AI Projects
Getting Started & Feasibility
Is AI right for my company?
AI is worthwhile when you want to automate recurring tasks, make knowledge accessible, or improve data-driven decisions. Typical entry points are AI chatbots, knowledge bases or process automation. In the initial consultation, we assess if and where AI provides the greatest leverage – sometimes a simpler solution is the better choice.
Do I need an AI strategy first?
Not necessarily as a 100-page concept – but a clear goal saves budget and time. In a discovery workshop or through our AI implementation service, we clarify use case, data situation, and success criteria before development starts.
Do I need large amounts of data for AI?
Not necessarily. Modern LLMs like GPT-4 already bring broad language knowledge. For chatbots and text processing, your manuals and process documents are often enough. An RAG knowledge base uses exactly these sources – as in our machinery manufacturing reference.
What does an AI project cost?
A simple chatbot/FAQ bot: from EUR 15,000 plus VAT. An AI knowledge base with RAG: EUR 30,000–80,000 plus VAT. Complex ML projects: from EUR 50,000 plus VAT. Plus ongoing API costs – typically EUR 100–500/month. Compare the business case with our AI ROI calculator.
Which reference project fits my use case?
For internal expert knowledge, see our machinery knowledge base. For consumer apps with personalised AI, see Chop-E. On-device ML is demonstrated by the Lullio baby monitor app.
Want to explore AI potential?
In 30 minutes, we will determine if and how AI can help you.
Bjoern Groenewold – Managing Director
Technology & Data Privacy
Which AI technologies do you use?
Depending on the use case: OpenAI GPT-4/GPT-4o, Claude, or open-source models for on-premise solutions. For embeddings and RAG we use Pinecone, Weaviate, or pgvector – typical in knowledge base projects. Voice use cases run through AI phone bots; autonomous workflows through AI automation.
What is RAG and when does it pay off?
RAG connects a language model with your company knowledge base: relevant documents are retrieved via vector search and included in the answer. This pays off when answers must be accurate, current, and traceable – e.g. for service manuals or internal policies. See our AI knowledge base service and the machinery case study.
Is this GDPR-compliant?
Yes – privacy is a project foundation, not an afterthought. Options include OpenAI with EU processing, Azure OpenAI in German data centres, or fully local open-source models. We align architecture with your industry – supplemented by GDPR-compliant software development and, where needed, guidance aligned with the EU AI Act as part of our AI services.
Can our data be used for training?
No, not with proper configuration. OpenAI API and Azure OpenAI do not use your data for training. For particularly sensitive data, we recommend local models that run entirely in your infrastructure.
How do you handle the EU AI Act?
We classify use cases by risk level, document data provenance, and add human-in-the-loop where transparency is required. For chatbots we check labelling obligations and technical safeguards against hallucinations – e.g. RAG with source citation, as part of our AI consulting and implementation
Implementation & Integration
How long does an AI project take?
An MVP chatbot: 4–6 weeks. A complete AI knowledge base: 2–4 months. Complex ML projects: 3–6 months. We start with a small proof of concept – often as an MVP to show results quickly before scaling.
Can AI be integrated into existing systems?
Yes – that is often the core benefit. We integrate AI into your website, CRM, ERP, intranet, or as a standalone solution. APIs connect to SAP, Odoo or custom software – built on our API and integration development.
Can AI be used on the phone?
Yes. AI phone bots handle first inquiries, appointment booking, or status requests – 24/7 – and connect to CRM, calendar, and ticket systems.
Who maintains the AI after launch?
We offer maintenance contracts for continuous improvement: updating the knowledge base, optimising prompts, incorporating user feedback. Alternatively, we train your team in our AI training programmes.
References & Success Measurement
How do I measure AI project success?
Define measurable KPIs before you start: e.g. reduced handling time in support, share of automatically answered inquiries, or time saved searching documentation. We document baseline and targets in discovery. Compare economic benefit with the AI ROI calculator.
Which industries do your AI references cover?
Projects range from manufacturing and machinery to consumer apps and IoT with on-device ML. Many AI features run in mobile apps.
Ready for AI?
Let us talk about your AI idea and plan the next steps.
Thorsten Frieling – Project Management