/
Advanced Data Analysis

Advanced Data Analysis

Owned by Marcin Sikora
Jan 13, 2025
4 min read

The Personal Assistant offers advanced data analysis features that empower users to gain insights from data files. Users can easily upload files, perform analyses, generate visualizations, and create documents based on their data.

Key Features

  • Interactive Exploration: The assistant allows users to ask follow-up questions and conduct iterative analyses, enabling deeper investigation of their data.

  • Visualizations: The Personal Assistant generates visualizations, such as graphs, to present data in an appealing and easily understandable format.

  • Document Generation: Users can create professional documents directly from their analyzed data, including presentations and reports.

  • Instructions Support: The user can provide specific instructions for data analysis, resulting in tailored insights and outputs.

Supported File Types

The Personal Assistant supports a range of file types, including:

  • CSV files (.csv)

  • Excel spreadsheets (.xls, .xlsx)

Getting Started with Data Analysis

To start using the data analysis features of the Personal Assistant, follow these steps:

  1. Enable Prompt Assistance:

    • Click on the "Options" cog icon located in the chat box.

    • Turn on the "Prompt assistance" toggle to activate the data analysis capabilities.

      image-20250110-133015.png

       

  2. Upload Files:

    • Click the paperclip icon in the chat message box to open the file upload dialog.

    • Select the desired file to upload.

    • The uploaded file will appear in the chat conversation, ready for analysis.

    • Uploaded documents remain available for conversation context during for 24 hours. After this period, you'll need to re-upload the documents to discuss or ask new questions about them.

      image-20250110-133324.png

       

  3. Perform Analysis:

    • Regular Data Analysis:

      • Ask questions related to data analysis, such as "Analyze sales trends in this dataset"

      • The assistant will process the data and provide detailed insights and summaries.

    • Visualizations:

      • Request visualizations by asking questions like "Create a bar chart of sales by region"

      • The assistant will generate interactive visualizations to represent the data.

    • Document Generation:

      • Instruct the assistant to create documents based on the analyzed data, such as "Create a PowerPoint presentation from this sales data"

      • The assistant will generate documents with the key findings and insights. You will be able to download a document. The unique link for the document will be valid for the next 24 hours.

  4. Data Analysis Responses

    • When the Personal Assistant processes a data analysis request, you will notice that the response indicates the specific data analysis AI model used, such as "GPT-4o-mini Data Analysis". This helps you distinguish data analysis responses from regular conversational responses.

       

  5. Iterative Analysis:

    • Engage in follow-up conversations with the assistant to refine your analysis.

    • Ask clarifying questions, request different views of the data, modify visualizations, or generate alternative document formats.

    • The assistant will maintain the context of the analysis session, allowing for seamless iteration and exploration.

Limitations

  • Personal Assistant can handle one data analysis request at a time. If you have multiple data files or analysis requirements, it is recommended that you process them sequentially. Running multiple analyses across different browser tabs is not supported.

  • Links to the generated documents are valid for the next 24 hours.

  • Uploaded files are stored for the 24 hours, after that period you should reupload them in order to refer to them in conversation.

  • If a response reaches the system's complexity limit, the assistant will pause and notify you. You can request the assistant to continue the response if needed. To prevent this, consider using prompts focused on a single analysis (ex. asking for one chart per prompt).

  • To initiate a data analysis request, the prompt must be classified as "Data Analysis" by the Prompt Assistance feature. For accurate classification, include a sentence that explicitly references data analysis tasks, such as statistical computations, trend identification, or insights extraction. Example: “Analyze the provided sales data to identify trends and generate a summary of key performance metrics.”

Examples

Here are a few examples of how you can interact with the Personal Assistant for data analysis:

User: Analyze sales trends in this dataset

Assistant: Let's perform these analyses step by step.

[Detailed analysis results showing trends over time]

 

 

User: Create a pie chart that represents the sales data by property type. Use a blue color palette.

Assistant: Here is the pie chart representing the sales data distribution by property type..

[Pie chart of sales data by property type]

 

 

User: Generate a powerpoint presentation with insights from data analysis of attached file

Assistant: The PowerPoint presentation on sales trends has been successfully created. You can download it using the link below:

[Download Sales Trend Analysis Presentation](Sales_Trend_Analysis_Presentation.pptx)