Capacity is missing for this task: Define tasks, output formats and test examples
Possible approach
Define tasks, output formats and test examples.
Specialist role
Prompt Engineer
You receive a versioned prompt and evaluation baseline. Define tasks, output formats and test examples. Compare prompt variants against consistent cases.
Understand the role
What does a Prompt Engineer do?
Define tasks, output formats and test examples. Compare prompt variants against consistent cases.
The central objective is: You receive a versioned prompt and evaluation baseline.
Problem → approach
Typical situations where this role helps
An AI demo looks convincing, but its quality on difficult inputs and its limitations are unknown.
Before a change, your team needs to address: Compare prompt variants against consistent cases
Possible approach
Compare prompt variants against consistent cases.
Your team needs a tangible output: A versioned prompt and evaluation baseline
Possible approach
Review failure cases and human oversight.
Does this fit your situation?Five short answers turn an initial idea into a first brief.
Check the fit ↗Inside the work
From problem to a verifiable outcome
An illustrative workflow for a Prompt Engineer. Select a step to see what may be prepared and handed over.
Starting point
An AI demo looks convincing, but its quality on difficult inputs and its limitations are unknown.
- Define inputs and evaluation criteria.
- Relevant systems: Python, OpenAI API.
Delivery
Define tasks, output formats and test examples
- Define tasks, output formats and test examples.
- Compare prompt variants against consistent cases.
Review
Review the work against a concrete situation
- Explain a representative evaluation set including difficult counterexamples and assess failures separately from averages.
- Also review for the assignment: Define tasks, output formats and test examples; Compare prompt variants against consistent cases; Experience with Python, OpenAI API.
Handover
A versioned prompt and evaluation baseline.
- A versioned prompt and evaluation baseline.
- Evaluated prototype with usage limits.
Typical projects
What an assignment could look like
Illustrative scenarios for orientation. Scope and outcomes are agreed for each assignment.
Define tasks, output formats and test examples
- Starting point
- Capacity is missing for this task: Define tasks, output formats and test examples.
- Approach
- Define tasks, output formats and test examples.
- Possible outcome
- A versioned prompt and evaluation baseline.
Compare prompt variants against consistent cases
- Starting point
- Before a change, your team needs to address: Compare prompt variants against consistent cases.
- Approach
- Compare prompt variants against consistent cases.
- Possible outcome
- Evaluated prototype with usage limits.
Handover for Prompt Engineer
- Starting point
- Your team needs a tangible output: A versioned prompt and evaluation baseline.
- Approach
- Review failure cases and human oversight.
- Possible outcome
- A documented working approach for Prompt Engineer.
Tangible deliverables
What may be delivered
Examples, not a blanket delivery promise. Choose the outputs your project actually needs.
- A versioned prompt and evaluation baseline.
- Evaluated prototype with usage limits.
- Review record for: Define tasks, output formats and test examples.
- Documented decisions, dependencies and open issues.
- Handover materials and knowledge transfer for the internal team.
Specialist fit
How to recognise relevant experience
For a Prompt Engineer, a traceable working approach matters. With VB Analyst, your task becomes a search brief with verifiable essential criteria.
Make experience tangible
Explain a representative evaluation set including difficult counterexamples and assess failures separately from averages.
- Connection to your assignment
- Define tasks, output formats and test examples
- Relevant working environment
- Python, OpenAI API
Anonymised examples suffice for an initial assessment. References, qualifications and availability are clarified for the assignment; a tool list alone does not establish suitability.
Which seniority makes sense?
An experienced specialist fits a well-defined package. Senior or lead experience matters more when the approach, interfaces or acceptance remain unclear. A junior profile needs a named specialist reviewer.
Applied to: Define tasks, output formats and test examples.
Remote, hybrid or on-site?
Remote work is usually practical with approved access, data and contacts. On-site sessions can support kick-off or handover.
A point to resolve in the brief
An AI demo looks convincing, but its quality on difficult inputs and its limitations are unknown.
Career profile · concise
Responsibilities, entry routes and working environment
For reference and preparation of your search brief.
Fact sheet: Prompt EngineerTasks · qualifications · tools
What does a Prompt Engineer do?
Define tasks, output formats and test examples. Compare prompt variants against consistent cases.
Tasks and responsibilities: Prompt Engineer
- Define tasks, output formats and test examples
- Compare prompt variants against consistent cases
How to recognise the outcome
A versioned prompt and evaluation baseline.
Training and degree paths: Prompt Engineer
Computer science, software engineering or electrical engineering; alternatively relevant vocational IT or electronics training with demonstrated development experience.
These are possible professional routes, not a universal degree requirement. For this role we review experience with a comparable task, technical depth and the ability to document a handover. Required degrees and evidence are defined in the specific search brief.
Specific selection questions
- Define tasks, output formats and test examples
- Compare prompt variants against consistent cases
- Experience with Python, OpenAI API
Tools & technologies
Python ↗Program data processing and reproducible analysis.OpenAI API ↗Integrate AI capabilities with defined inputs and evaluation.
Possible working environment; the actual combination depends on the assignment.
Capability compass
Which combination moves your project forward?
Connect your task to relevant capabilities. A tool selection narrows the working environment; the results explain each professional connection.
Starting pointPrompt EngineerSearch the full catalogue ↗
03Reasoned role selectionTask · capabilities · outcome
The professional connection becomes clear through tasks and possible outputs.
Data scientist
Explore data and develop suitable statistical models for a specific question.
Your possible outcome
Analysis with a baseline, validation and clear interpretation of limitations.
Machine learning engineer
Turn a validated model into a maintainable technical workflow.
Your possible outcome
Versioned model deployment with monitoring, tests and a documented fallback.
MLOps engineer
Connect model versions, data dependencies, deployment and monitoring in a reproducible workflow.
Your possible outcome
Versioned model pipeline with checks, operational metrics and documented rollback.
Capability profiles for orientation. An individual’s suitability is assessed against the search brief.
Refine the selection ↗Define the boundaries
When another role may fit better
This may not be the right role if your main priority lies elsewhere. These profiles help clarify the difference.
If your priority is: Connect models, data access and application components. Design operations, access boundaries and quality checks.AI Solution Architect ↗If your priority is: Implement model calls and application interfaces. Test quality, latency and failure behaviour in context.LLM Engineer ↗If your priority is: Structure document access and retrieval. Test retrieval and answers against verifiable reference cases.RAG Engineer ↗
Roles compared directly
This overview describes typical areas of responsibility. Actual scope may vary between organisations.
| Criterion | Prompt Engineer | AI Solution Architect | LLM Engineer | RAG Engineer |
|---|---|---|---|---|
| Core task | Define tasks, output formats and test examples. Compare prompt variants against consistent cases. | Connect models, data access and application components. Design operations, access boundaries and quality checks. | Implement model calls and application interfaces. Test quality, latency and failure behaviour in context. | Structure document access and retrieval. Test retrieval and answers against verifiable reference cases. |
| Possible outcome | A versioned prompt and evaluation baseline. | An AI solution design with documented decisions. | An evaluated language-model feature with a clear operational handover. | A retrieval pipeline with source references and documented evaluation limits. |
| Working environment | Python, OpenAI API | Python, Microsoft Azure, OpenAI API | Python, PyTorch, OpenAI API | Python, PostgreSQL, OpenAI API |
Unsure which role fits?Start with your goal and your team’s tasks.
Start the role finder ↗Divide the work sensibly
Which expertise complements this role?
Complementary roles address adjacent tasks. They are not automatic substitutes for a Prompt Engineer.
Responsible AI Consultant
Specify AI quality and responsibility questions. Align evaluation and escalation rules with teams.
Agree the interface
Discuss this combination ↗A use-case-specific responsible AI design with test cases.
Data engineer
Connect data sources and develop traceable processing pipelines.
Agree the interface
Discuss this combination ↗Versioned data flow with quality rules, exception logging and operational handover.
Which work can be scoped as a package?
A managed service requires defined inputs, scope and approval paths. These services provide a starting point for that definition.
AI & machine learning
A scoped modelling or analysis brief with review criteria and documented limitations.
Choose work packages ↗Workflows & AI assistance
A testable workflow with defined inputs, human approvals and exception handling.
Choose work packages ↗For agencies and service providers: White-label delivery can align formats, approvals and communication under your brand. Client access and responsibilities are agreed in advance.
Interactive fit check
Does a Prompt Engineer fit your project?
Five questions, a reasoned assessment and a brief for your enquiry. You can change every answer.
Your assignment with VB Analyst
Choose expertise. Define the engagement.
A capacity gap does not always require a permanent role. Choose a model by responsibility, duration and desired outcome.
Useful information for your enquiry
Prompt Engineer
Which errors are unacceptable, which data may be used and who evaluates responses?
Possible work objective
A versioned prompt and evaluation baseline.
- Task, existing systems and scope
- Essential criteria, internal contact and approval
- Preferred start, duration, capacity and budget range
You can leave undecided details open. Non-confidential information is enough for initial contact.
Selected model: Project support
Discuss these requirements ↗View this model and its responsibilities ↗What happens next
From enquiry to a well-prepared start
- Define the requirement
We clarify the task, priority and outstanding requirements with you.
- Assess specialist fit
Relevant experience is assessed against the assignment. Open questions and working parameters remain visible.
- Agree selection and scope
You decide through specialist discussions. Capacity, terms and responsibilities are agreed.
- Prepare onboarding
Access, the first milestone, contacts and handover are established.
How soon can work start?
Timing depends on suitable availability, selection, agreement and access. For urgent needs, separate essential initial work from later tasks. A binding start date is confirmed for the specific assignment.
A useful starting point
Anything still unclear?
Short answers for your next step. We can work through your specific situation together.
Discuss my question ↗What does a Prompt Engineer actually do?
Define tasks, output formats and test examples. Compare prompt variants against consistent cases. One possible outcome: A versioned prompt and evaluation baseline.
How can I assess professional fit?
Explain a representative evaluation set including difficult counterexamples and assess failures separately from averages.
Which tools does the specialist need?
Possible working environments include Python, OpenAI API. The required combination depends on your assignment. Not every listed tool is a mandatory requirement.
Are the specialists available now?
The profiles describe capabilities and typical assignments. Actual people, availability, terms and engagement are assessed for your specific need.