> ## Documentation Index > Fetch the complete documentation index at: https://labs.laer.ai/llms.txt > Use this file to discover all available pages before exploring further. # Chat prompt tips Creating effective prompts helps Ask AI deliver more accurate and relevant responses. Use the following tips to get the most out of your conversations. ## Provide specific details The more detail you include, the better Ask AI can understand your intent. Example
Instead of:Try:
Tell me about LJM2. What did Enron’s CFO Andrew Fastow say about LJM2’s structure?
## Maintain consistent context Stick to one topic for each conversation. Shifting topics can confuse Ask AI and reduce response quality. Start a new conversation when switching topics. ### Example
Instead of:Try:

What are the main issues and implications discussed regarding off-balance sheet entities in the Enron documentation?

and

Create a summary analyzing the insider trading conducted by Ken Lay and Andrew Fastow. Include supporting evidence with DOC ID citations instead of index numbers.

Start a new conversation for each topic to keep responses focused and accurate.

## Use quotes for exact search Use quotation marks to search for exact phrases. This ensures the results find the words used together in the specified order. ### Example
Instead of:Try:

Find examples of documents that mention insider trading.

This prompt returns results that mention the terms insider, trading, or insider trading.

Find examples of documents that mention “insider trading”.

This prompt only returns results that mention the term insider trading.

## Focus on topics or ideas Ask about ideas or issues rather than requesting lists of documents. This leads to more insightful responses. For instance, a more effective prompt might summarize key points from various documents, highlight significant discussions, and provide a clearer understanding of the topic. ### Example
Instead of:Try:

List all documents that mention off-balance sheet entities.

What are the main issues and implications discussed regarding off-balance sheet entities?

## Avoid number-based prompts Refrain from asking Ask AI for counts or quantities. It is better to focus on meaning over metrics. **Note:** For document counts or search metrics, use your eDiscovery platform directly. ### Example
Instead of:Try:

How many documents mention off-balance sheet entities?

What are the implications of off-balance sheet entities as discussed in the documents?

## Rephrase when needed If a prompt does not return the expected results, try rephrasing, breaking it into parts, or using synonyms. ### Example
Instead of:Try:

Was this a violation of GAAP?

Did the documents discuss whether the accounting followed GAAP rules?

## Build on results Use snippets and citations from responses to refine your prompts or ask follow-up questions. ### Example
Instead of:Try:

List all documents that mention off-balance sheet entities.

Since this prompt centers on a list, there is no deeper insight to expand upon.

What are the main issues and implications discussed regarding off-balance sheet entities?

Then:

Can you provide more details about how Enron’s off-balance sheet entities were structured?