AI Glossary

A company cockpit maps a company's structure and brings phone, e-mail, inbox and forms together in one place. In it, AI agents take over recurring tasks such as inquiries, quotes, receipts, appointments and knowledge questions; the team sees everything at a glance, steers the agents and approves. mysoftwarelab builds the cockpit individually and runs it as an external AI department.

An AI agent is an AI that handles a task end to end on its own – e.g. capturing a customer inquiry, preparing a quote or matching a receipt – instead of just generating text. In the company cockpit, agents work with clear tasks, hand-offs and approvals: agents take over tasks, not people.

The AI maturity level describes how far a company has progressed in adopting and using artificial intelligence. Typically 5 stages: (1) Awareness, (2) Experiments, (3) Pilots, (4) Scaling, (5) AI-native company. The maturity level determines the right entry strategy.

An AI roadmap is a structured timeline for the step-by-step introduction of AI solutions in a company. It includes use case prioritization, resource planning, milestones and KPIs. A typical AI roadmap for SMEs has a horizon of 12–24 months.

SMEs are companies with fewer than 250 employees and annual revenue under €50 million (EU definition). In Germany, around 99% of all companies belong to the Mittelstand. mysoftwarelab works with companies from 20 employees: there, AI agents in the company cockpit take over recurring tasks.

Prompt engineering is the art of formulating instructions (prompts) for AI systems such as ChatGPT or Claude so that they consistently deliver high-quality results. Well-designed prompts can solve 80% of quality issues with AI outputs. A key competency for AI adoption.

LLMs are AI models trained on huge amounts of text that can understand and generate natural language. Well-known LLMs: GPT-4 (OpenAI), Claude (Anthropic), Gemini (Google). They are the technical foundation for chatbots, text creation, code generation and many automation processes.

Generative AI refers to AI systems that can create new content — text, images, audio, video or code. In a business context, it is used for text automation, image generation, customer communication and software development. Difference from classical AI: creates instead of only analyzing.

An AI use case is a specific business process that can be improved or automated through AI. Examples: automatic invoice processing, AI customer support, predictive maintenance, AI-supported quote generation. Identifying suitable use cases is the first step in any AI adoption.

ROI measures the ratio of profit to investment costs. Typical for AI projects: initial investment (consulting + implementation) vs. ongoing savings (time savings, error reduction, higher conversion). Our clients typically achieve 60–80% cost savings in automated areas, with ROI realized after 6–14 months.

An AI pilot project is a time-limited, controlled first test of an AI solution in a company. Goal: prove feasibility, measure ROI and test acceptance — before full scaling begins. Typical duration: 4–8 weeks. Important: define clear success criteria.

Change management describes all measures that accompany employees during the introduction of new technologies. Particularly important for AI projects: reducing fears ("AI will take my job"), training, clear communication and early involvement. 70% of all AI projects fail not because of technology, but because of change management.

Process automation means replacing recurring manual tasks with software or AI. AI automation goes beyond classical RPA: it can process unstructured data, understand natural language and make context-dependent decisions. Typical savings: 60–80% of manual processing time.

The EU AI Act is the world's first comprehensive AI regulation, which entered into force in August 2024. It classifies AI systems by risk level (unacceptable, high, limited, minimal) and prescribes requirements for transparency, documentation and human oversight. Relevant for SMEs from 2025.

AI governance comprises all rules, processes and structures that ensure the responsible use of AI in the company. This includes: data quality assurance, bias detection, audit trails, compliance with the EU AI Act and internal AI guidelines. The foundation for trustworthy AI.

Predictive maintenance uses AI and sensor data to predict machine failures before they occur. Typical in manufacturing: reduction of unplanned downtime by 40–60%, extension of machine lifespan by 20–30%. Especially relevant for mechanical engineering SMEs.

RAG is an AI architecture that connects LLMs with your own knowledge base. The result: an AI system that answers based on your own documents, manuals or product data — without hallucinations. Frequently used for internal knowledge bases and customer chatbots.