Video walkthrough — plays on the live article.

With Reporting in Aline, you can:

This makes it easy to move from document-level review to portfolio-level insights.

To analyze multiple agreements at once:

Once inside a report, you'll see all documents currently stored in your repository.

You do not have to build a report column by column. Click the + in AI Reports and choose Create report with AI, then describe the report you want — for example, a report of all my vendor agreements showing payment terms, renewal type and governing law. You can attach a sample agreement so it knows what the documents look like.

The AI chooses the prompts, the input and the output type for each property, so the columns arrive ready to run. You can do the same thing from the AI chat — create a report for all my NDAs and extract provisions around trade secrets.

⚠️ If you name a document type in your description, the report is also filtered to that type. Asking for "vendor agreements" analyses everything but shows you only vendor agreements. That is usually what you want — just be aware the filter is there if a document you expected is missing.

⚠️ The chat can create a report but cannot edit one. Asking it to change an existing report produces a second report instead. To change a report you already have, open it and use Filters → Add filter, then save your changes.

💡 A report is a point-in-time view of a question. Rather than deleting one and rebuilding it, keep it and create a new report for the new question.

⚠️ Before you click run, decide what you are running on. If nothing is selected, the property runs on every document in the report — which on a large repository can be thousands of files and a large token spend. To run on a subset, select those documents first, then run.

If you click "Run Aline AI" with no selections when you already ran the AI for a few documents, the documents with the responses will not be rerun. Run Aline AI only runs on whatever it has not run on already.

See Understanding Token Usage in Aline.

To narrow your analysis to specific agreements:

💡 Best Practice: Use labels as your primary way to categorize documents — this makes filtering in reports fast and reliable. Add all the columns you want before you filter, so the AI runs across the full set.

📌 A property cannot be scoped to a single document. If you only want to run a property on a handful of files, filter the report down to those files, run the AI, then remove the filter.

Once your document set is selected, you can extract insights in two ways.

Option A: Use an Existing Prompt

Option B: Create a New Property

With single-select and multi-select you give the AI the list of allowed options, which keeps answers consistent and makes grouping and filtering reliable. For guidance on which type to pick, see Choosing the Right Output Type for Report Properties.

📌 Number columns sort by value, so a contract-value column orders properly rather than alphabetically.

⚠️ The Number output does not accept currency symbols or thousands separators. For a figure like total contract value, extract the value as Text if you need it to read exactly as written, and add a second Number property holding the bare figure if you need to sort or filter by range.

📌 A Date property on an evergreen or auto-renewing contract comes back as dashes, because there is no fixed end date to extract. See Managing Contract Status, Expiration & Auto-Renewal (Evergreen) Dates for how to give those agreements a real date to remind against.

To change a property's prompt after you have created it, hover over the property and click the pencil icon. Edit the wording, then re-run.

You do not have to re-run a property to fix one wrong answer. Click the cell and type the correct value. The cell is marked as edited by you, so anyone reading the report can tell your correction apart from an AI answer.

⚠️ A column you fill in by hand cannot then be used as the input to an AI column — AI properties read the file or another AI-extracted property.

⚠️ Testing ensures accuracy and saves tokens when running large-scale reports.

This makes it easy to verify findings and drill into specific agreements.

📌 Everything a report extracts is saved onto the document itself, so the data is available outside the report too.

Once a report exists, the AI chat can read it and give you aggregate answers rather than rows — for example, how many of our contracts expire after this date? returns the number, with links to the documents behind it.

💡 Name the report in your prompt so the chat scopes the answer to it. The chat can also answer across your whole repository, and will tell you if some contracts have not been analysed yet.

📌 Charts generated from a report are not available yet — export the data if you need to visualise it.

By default a property reads the file. You can instead point it at a column you have already extracted, by changing the property's input — useful when you want the AI to classify or judge an answer rather than re-read the whole contract.

For example:

A useful pattern is to extract the clause verbatim in one column, then classify it in a second column that reads the first.

You can export Aline reports as a CSV or Excel file by clicking ACTIONS at the top left and selecting the file type you'd like to export.

Click the Support button in Aline to reach our team.