Usage guide for Simien AI-assistant

Good examples from real usage:

• “Give me step by step how heat transfer is modelled in NS 3031”
• “Can you give me a detailed table for every energy rating vs energy consumption?”

Guideline:

Start your question with what you want the chatbot to produce:

• an explanation
• a table
• a comparison
• step-by-step instructions
• implications or consequences

High-quality questions explicitly reference standards, regulations, or tools.

Good examples:

• “…modelled in NS 3031”
• “new energy rating from 1st January”
• “according to TEK17 and NS 3031:2025”
• “How is this handled in Simien PRO?”

Guideline:

Whenever relevant, explicitly name:

• standards (NS 3031:2025, TEK17, …)
• software (Simien PRO, Simien 6)
• regulation versions or dates

The best questions clearly define how broad or narrow the answer should be.

Good examples:

• “Give me a detailed table for every rating vs energy consumption”
• “Explain in detail what EPB is and its implications”
• “Give me the energy labeling table (A–G) for barnehage according to NS 3031:2025.”

Guideline:

Explicitly say if you want:

• a high-level overview
• a technical explanation
• tables, formulas, or examples
• a single building type or all categories

The chatbot is designed to work conversationally. You can ask follow-up questions to:

• Clarify parts of an answer
• Narrow down options
• Request examples, explanations, or alternative approaches

Good example flow:

1. “What can you tell me about the new grading scale?”
2. “Can you give me a detailed table?”
3. “Give me the table for barnehage.”

Guideline:

Use the session to:

• narrow from general → specific
• request clarification or missing pieces
• zoom in on one building type, scenario, or assumption

💡 You can ask up to 10 messages in one session, so use these to progressively improve and refine the response.

❗Sometimes the answer might stop midway. In this case, just text “Come again” or “Repeat that” to regenerate the complete answer OR start a new conversation

The best technical questions include conditions that affect the answer.

Good examples:

• specifying energy labeling vs TEK compliance
• referencing normative vs measured data

Guideline:

Include relevant constraints such as:

• building type or use
• regulatory purpose (energy label, TEK check, analysis)
• timeframe (new rules, upcoming changes)
• whether you want normative or realistic assumptions

If something looks incorrect or unclear, point it out and explain what seems off. The chatbot can adjust its response when given additional guidance or corrections.

➡️ In general: The more relevant context and guidance you provide, the more accurate, useful, and tailored the chatbot’s response will be.