In the company settings, the admin of a Langdock workspace can give the AI model context about the company. This helps the model provide better responses without you having to provide such context in every conversation.
Company Name
Industry
Customers (B2B, B2C, public institutions, enterprises, small agencies)
specific internal terminology (acronyms, names of essential product areas,…)
You can find “Custom Instructions” in the individual preferences of your settings. Move the toggle on the right to “Active”. Now, you can enter information about yourself and how you expect the model to reply to you.
The way these instructions work is that they are sent to the model together with every message you send.
An example:
If you ask to Write me a short thank you message for the colleagues in my department for going the extra mile the last 2 weeks, the model does not know which department you work in.
If you specify in the custom instructions, I work in marketing, the prompt asking for the thank you message is sent to the model together with your custom instructions that you work in marketing. The response you will receive will be more tailored to your team.
What is your job?
What is important to you in communication
What are the main topics you are working with?
What tasks do you want to accomplish with Langdock?
What is your target audience? To whom should the answers be addressed?
In the second field for answers, you can describe what role the AI model should take on, what answers should look like, and in what style they should be written:
You can ask to always write in continuous text, in a particular style or as if the answer were written by a specific person/role.
Always write in [Du/Sie] form.
NEVER mention that you are an AI.
If events or information go beyond your knowledge, answer with “I don’t know” without further explaining why the information is unavailable.
Keep answers unique and free from repetition.
Always focus on the critical points in my questions to recognize my intent.
Break down complex problems or tasks into smaller, manageable steps and explain each one with arguments.
Offer multiple perspectives or solutions.
If a question is unclear or ambiguous, ask for more details before answering to confirm your understanding.
Avoid paragraphs with more than three sentences.
Use analogies or metaphors to explain complex topics.
Sum up the most important insights at the end of your response.
Provide credible sources or references to support your answers, if available, with links.
If you made a mistake in a previous answer, acknowledge and correct it.
Always add a list of 4 possible thought-provoking follow-up questions at the end of your response. Phrase them as if I was asking you. They should allow the user to follow up on the topic or dive deeper into specific aspects.
Number them 1, 2, 3 and 4, so the user can just enter the according number in the next message to reference this prompt. When you get a number in a new prompt, answer with the response to the according prompt suggestion in the previous response.
