To customize any list with enriched data, open your list and click the +AI Column button at the top-left of your list to create analytic columns.
The AI Column feature is an integral tool for list-building workflows, which are frequently leveraged to enrich and accelerate insights.
An overlay titled Table Blueprint Preview opens.
There are 3 ways one can create AI Columns:
Option 1:
Generate Blueprint functionality is effective in crafting highly customized, purpose-built tables that align with specific analytical or reporting needs.
In the center text box, describe what you need and the AI will generate a systematic blueprint from your instructions.
Enter your query and click Submit.
Based on your input, you’ll receive a few clarification questions to better understand the scope of the analysis before generating the columns. If you realize you missed something in your initial description, feel free to include it in your scope response.
Type your answers, click Submit and wait for the Blueprint Preview.
Option 2:
Click the grid icon in the top-right of the overlay to open the Column Configuration panel. Then click the + next to Column Configuration to manually add columns.
You can:
Add/remove columns
Update column titles and descriptions
Rearrange or duplicate columns/sub-columns
Choose between:
Single Columns: Use this if you just need one piece of data per row (e.g., “Dosing Guidelines”)
Grouped Columns This powerful feature allows you to efficiently gather multiple related data points into separate sub columns from a single AI-driven research pass, much like how you'd typically populate a detailed spreadsheet. This capability dramatically speeds up your research, providing richer insights without extra effort.
For example, you can simultaneously retrieve multiple clinical trial outcomes such as Overall Survival (OS), Progression-Free Survival (PFS), details about the active and placebo arms, and the specific measurement timeframe, all clearly organized into separate but related columns.
Option 3:
You can also click the book icon in the top-left corner to open the Blueprint Library.
Blueprint Library: Explore Maven Bio’s prebuilt Blueprints for common analysis types. Use dropdown menus to filter, search by keyword, or navigate with on-page commands.
If any built-in blueprint is relevant to your analysis, you can select it and click Apply to use it. You can also click the star icon to add the blueprint to your Starred Blueprints list for easy access in future projects.
To save your blueprint, click on the "Save Blueprint" button- enter a name, choose the row type (e.g., trial, drug, company), and add an optional description for context.
Then click Update Blueprint to apply changes to the current blueprint, or Create New Blueprint to save a new one.
Regardless of how you build your blueprint, a live preview appears in the center of the overlay. From this preview, you can edit or delete columns directly.
Hover over any column header or cell to view its details.
If everything looks correct and aligns with your analysis goals, click Apply Blueprint to generate your final table with enriched, AI-powered data.
Once the blueprint is applied, the relevant columns are automatically initialized, and Maven’s research agent begins its analysis based on the criteria defined in the selected blueprint.
You’ll see new columns added to the list, and a pop-up notification will appear once the research is complete.
Hovering your cursor over the purple line atop the column headers reveals the column descriptions you provided.
If you hover directly over a column title, you’ll see the specific instructions that Maven’s research agent is following to generate responses for that column.
Users can delete any column from the table to tailor the view to their needs. Additionally, if the cell data supports it, columns can be sorted and filtered, giving users greater flexibility in analyzing and customizing their datasets.
Columns added to the table, whether via Blueprint or manually are listed under the Columns tab on the top left of the screen. If users prefer not to delete a column but want to focus on specific fields, they can use the checkbox controls to hide or unhide columns as needed. This makes it easy to customize the view without losing any underlying data.
Each cell in the table is expandable and collapsible, allowing for flexible review. Cells researched by the agent include a colored dot in the top-right corner, indicating the strength of supporting evidence:
🟢 Green- Direct evidence
🟡 Yellow- Indirect evidence
🔴 Red- Minimal evidence
Clicking on a cell reveals a detailed view with direct links to key supporting evidence, the full research log, and the sources used to formulate the response. This expanded view can also be opened in a separate window for easier navigation and comparison.
The research log offers a step-by-step breakdown of Maven’s workflow, tracing the process from initial prompt handling and query analysis to resource review and final synthesis of the response.
Users always have the option to export a table as an Excel or CSV file for offline access and analysis. Even after export, each file retains direct links to the evidence log for each cell, allowing users to revisit original sources or review the reasoning behind any cell’s response. This ensures traceability, transparency, and continuity across workflows, even outside the platform