An AI team is a cross-functional group combining AI, data, software, UX, product, and domain expertise to develop an AI product. Its members may use different goals and vocabulary, so collaboration must align technical, domain, and user concerns around shared outcomes. section 12.1 section 12.1.1
Effective AI-team collaboration depends on three practices:
- Understand different roles, backgrounds, tasks, and pain points so conversations create less friction. section 12.1
- Replace isolated role metrics with shared user-centered goals such as satisfaction, adoption, and practical usability. section 12.1
- Iterate and learn together because AI development involves uncertainty, experimentation, testing, and collective learning. section 12.1
- Choose team structure contextually: roles may be unnecessary, stage-specific, or reshaped as development moves from model selection toward scalable infrastructure. section 12.1.1
Accuracy is not the whole product
A data scientist celebrates a 4% improvement in recommendation accuracy, while a UX designer notes that users still do not trust the product. The team must connect model performance to adoption and practical usability rather than treating accuracy as the sole success measure. section 12.1
Table 12.1 Five core areas of competence in an AI team
| Area |
Roles |
Description |
|---|---|---|
| Product management |
Product manager |
Identifies the opportunities, customer needs, and larger business objectives that the AI system will fulfill. The project manager also communicates the requirements and success criteria and guides a team in implementing this vision. |
| Software engineering |
Backend engineer Frontend engineer DevOps engineer Software architect |
Focuses on building, maintaining, and scaling the technical infrastructure of AI systems. This includes backend, frontend, and DevOps engineers, who handle the software foundation, and roles such as software architects, who design the overarching system structure. |
| Data science and AI development |
AI engineer Machine learning engineer Data scientist Data engineer MLOps engineer Prompt engineer |
Specializes in designing, training, and operationalizing AI models. This area includes machine learning engineers, data scientists, and MLOps engineers, as well as roles such as prompt engineers and AI engineers who integrate AI components into usable systems. |
| User experience |
UX researcher UX designer UI designer Conversational designer Content designer |
Ensures the AI system is user friendly, engaging, and aligned with user needs. UX and UI designers craft the functional and graphical experience, while conversational and content designers focus on interaction flows and clear communication. |
| Domain expertise |
Domain expert Data annotator |
Provides encoding-relevant expertise for the AI system. This includes the compilation and annotation of specialized datasets. |
Use the competence map to distinguish product management, software engineering, and AI-development responsibilities when shaping an AI team. section 12.1.1
