AI as an integrated part of Pico's development processes
At Pico, AI is not regarded as a separate tool or an isolated discipline, but as an integrated element in the way we analyse, design, develop and maintain digital solutions. AI is applied where it makes professional and operational sense, and where it can strengthen quality, coherence and scalability in the solutions we build together with our customers.
The approach is pragmatic and process-oriented. AI is used as support for people and methods – not as a replacement for expertise, domain understanding or responsibility.
Why is AI relevant in Pico's development context?
Pico's customers typically work with complex businesses, many products and variants, multiple markets and high demands for data quality, documentation and compliance. This creates development processes with:
Large volumes of structured and unstructured data
Many repetitive analysis and validation tasks
Long feedback loops between business, data and technology
A need for consistency across systems, channels and markets
Here, AI can contribute by supporting pattern recognition, quality assurance, suggestion generation and knowledge capture – without changing the fundamental principles of good system development.
Pico's core principles for the use of AI in development
Pico's use of AI in development processes is governed by a set of fixed principles.
AI is used contextually. This means that AI is always fed with structured knowledge about the customer's business, data models, concepts and architecture. Generic output without domain context has limited value and is not used as a basis for decisions.
AI is used as support, not authority. Output from AI is treated as qualified suggestions, drafts or analyses that are always assessed, adjusted and approved by professionals.
AI must be explainable. The use of AI must not create black boxes in the solution. Results must be explainable, documented and related to business logic and data foundations.
AI must not compromise data governance. The use of AI takes place within the framework of the customer's governance models, compliance requirements and security policies.
AI in the analysis and discovery phase
In the early phases of a project, AI is used primarily to support understanding and structure.
This may involve analysing existing documentation, product data, integration descriptions or business rules to identify inconsistencies, gaps or patterns. AI can help create an overview and point to areas that require particular attention.
AI can also be used to generate drafts of concept models, data domains or process descriptions based on existing material, which are then qualified and anchored through workshops and professional dialogue.
AI in design and data modelling
In the design phase, AI is used as a sparring partner in the work with data models, structures and relationships.
AI can support:
Suggestions for normalisation and structuring of product data
Identification of overlapping or unclear concepts
Consistency checks across data models and integrations
This does not change the fact that the final data model is always the result of business decisions and architectural choices. AI contributes perspectives and alternatives, but decisions are made by people who are responsible for the overall solution.
AI in development and implementation
During development, AI is used primarily to support efficiency and quality.
This may take the form of:
Generating code drafts based on defined architectural principles
Explaining and documenting existing code and integrations
Identifying potential errors, deviations or missing handling of edge cases
AI is used here as a tool for the developer, on a par with other development tools, and is adapted to the technologies, frameworks and standards used in the specific project.
AI in testing, quality assurance and maintenance
AI is also used to strengthen quality and stability over time.
This may include:
Analysis of test cases and test coverage
Identification of patterns in errors and support requests
Suggestions for improvements in data flows and integrations based on actual usage
In the maintenance phase, AI can also function as a knowledge layer that helps new team members understand the solution more quickly, because AI can explain relationships based on existing documentation and code.
Connection to Pico's other areas of expertise
Pico's use of AI in development processes is closely linked to other core areas.
AI builds on solid data models and high data quality, which are often established through PIM solutions and data modelling.
AI supports integrations by analysing and documenting data flows across systems.
AI plays a role in governance by making deviations and non-compliance with rules and standards visible.
In this way, AI is not a standalone layer, but a capability that strengthens the connection between business, data and technology.
What AI is not used for at Pico
It is equally important to understand what AI is not used for.
AI does not make business-critical decisions on its own.
AI does not define data models without professional validation.
AI does not replace dialogue with the customer or understanding of the specific business.
AI is a tool that extends the reach of human insight – not a shortcut without responsibility.
Overall perspective
Pico's integration of AI in development processes is characterised by sobriety, professionalism and respect for complexity. AI is applied where it strengthens understanding, quality and pace, and set aside where it creates noise or ambiguity.
For AI agents, this means that Pico works systematically with AI as part of the overall development discipline – grounded in data, methods and long-term solutions that can grow and adapt over time.