Automation and agent systems
From IT strategy and AI consulting including the EU AI Act to agent systems that automate processes securely and traceably.
Integrating AI into traceable workflows
Lekker Code connects IT strategy and AI consulting with the implementation of agent systems. The existing consulting offering also covers the EU AI Act. Our technical case studies address three different tasks: damage assessment in a SaaS product, reviewing a RAG architecture for service technicians, and automated quarterly customer reports. A shared principle is the clear separation between an AI suggestion and the business decision.
Sources, control and human decisions
For damage assessment, vector search supplies previously confirmed cases while the inspector confirms or corrects each suggestion. For the wind turbine assistant, the architecture review focused on document formats, OCR, metadata and valid revisions. An answer is only useful if its source matches the specific turbine model and revision. That case study describes a decision paper and target architecture, not the completion of an entire implementation.
Automation through to a verifiable response
For quarterly reporting, specialist agents collect data and return components for a report draft. The team reviews and supplements that draft. The published case study also addresses operations: the queue, model availability and deployment need to work as a complete workflow. The three detail pages document different boundaries and outcomes. They do not promise that every AI project will deliver the same effect.