What are Unique Reports?

Reports allows you to prompt GPT-3 to perform tasks.

OpenAI has trained cutting-edge language models that are very good at understanding and generating text. With our Reports feature, we’ve integrated the power of GPT-3 to make it easier for you to perform a wide range of tasks, such as:

  • Summarization (e.g. key information) for generating meeting minutes, a high-level overview or executive summary.

  • Content creation (i.e. generate text in various forms) for drafting follow-up emails, descriptions, and announcements.

  • Topic modeling (i.e. main topics or themes discussed during the meeting) for identifying key takeaways or areas of focus.

  • Entity recognition (i.e. people, organizations, and locations) for tracking the involvement or interest of different stakeholders.

  • Information recognition (i.e. names, dates, statistics, and other key data points) populating forms and records.

  • Action item identification (i.e. specific action items or to-dos that were discussed) for ensuring action items are not overlooked or forgotten.

These are just a couple of examples among many possible use cases.


How it works

1. Visit a meeting's recording page.

2. Scroll down to the Reports section.

3. Type your prompt.

Prompts are how you get GPT-3 to do what you want. So, you have to know what you're trying to accomplish.

4. Click the blue button to send your request.

Note: When a prompt is sent to GPT-3, it's broken down into tokens. This enables GPT-3 to work with relatively large amounts of text. It can take about 2 minutes or more to generate your Report.


Best Practices

Use clear and concise language

Make sure your prompts are easy to understand and provide clear instructions. Avoid using technical jargon or complex language.

Less effective ❌:

Generate a summary of the client meeting.

Better ✅:

Summarize the key investment recommendations discussed during the client meeting with Mr. Hauri on March 1st. The recommendations should include both short-term and long-term strategies, with an emphasis on risk management and tax efficiency.

Be specific, descriptive and as detailed as possible about the desired context, outcome, length, format, style, etc

Avoid general or open-ended prompts, as these can result in less useful or irrelevant responses. Instead, use specific prompts that are tailored to the task at hand.

Give GPT-3 enough context by providing relevant information about the meeting and its purpose. This will help GPT-3 generate more accurate and relevant responses.

Less effective ❌:

Write a summary of the meeting.

Better ✅:

Write a summary of the key discussion points. The sales team meeting was focused on discussing the Q4 sales targets and progress towards meeting those targets. Make sure the length is 5-10 sentences long. 

Articulate the desired output format through examples

Less effective ❌:

Extract the following 4 entity types: company names, people names, specific topics and themes.

Better ✅:

Extract the important entities mentioned in the text below. First extract all company names, then extract all people names, then extract specific topics which fit the content and finally extract general overarching themes

Desired format:
Company names: <comma_separated_list_of_company_names>
People names: -||-
Specific topics: -||-
General themes: -||-

Test and iterate

Continuously test and refine your prompts to ensure that they generate the desired results.


OpenAI Resources

Did this answer your question?