Map Coverage with the Feature Matrix

Once you've found relevant patents, you often need to answer a harder question: which specific features does each patent actually cover? The Feature Matrix lets you map elements against references and build a structured coverage picture — with AI filling in the evidence so you can focus on reviewing it.

When to use the Matrix

The Matrix fits any workflow where you need to compare what multiple patents cover:

Invalidity analysis — map claim elements against prior art to show which references read on which limitations
Freedom-to-operate — check whether your product features appear in patent claims
Novelty assessment — identify which elements of your invention are already covered and where the gaps are

Most users use it as the final step after a search to structure, consider, and report what they've found. Other users who already have some patents on hand may start directly with coverage analysis by uploading those patents and running the AI even without doing a search first.

Step 1: Define your elements

Elements are the rows of the Matrix — claim limitations, product features, or invention elements depending on your project.

Click + Add Elements and choose Generate with AI. Amplified reads your project's description and claims fields to propose a set of elements. You'll see each one with a checkbox, a short label, and a more detailed descrition. Edit any element inline, uncheck the ones you don't need, and click Add to matrix.

This allows you to quickly turn products, claims, or inventions into a clear set of distinct features. The matrix will use these later to map features to specific evidence in each patent you add.

If the features aren't exactly what you want, you can either give the AI guidance and regenerate or add elements manually.

Tip: If you expect precise matches, use precise language — "capacitive touch sensor on the display surface" will get very different AI coverage results than "touch screen."

Step 2: Add references

References are the columns — the patents you're evaluating.

Click Add Reference and you'll see patents you've marked as relevant during your search. Select the ones you want to analyze and click Add to matrix. You can also add patents manually by number, or add non-patent literature (coming soon).

Step 3: Let AI fill coverage

Click Fill coverage in the toolbar. For every cell in the matrix — each element-reference combination — AI reads the patent text and:

Extracts the supporting passages with source locations
Sets a coverage level: Yes, Maybe, No, or Unknown
Writes reasoning explaining why

This is the step that replaces hours of manual reading.

Step 4: Review and refine

AI-filled coverage is a starting point. Your job is to verify it.

Click any cell to open the detail panel. You'll see the coverage level, the AI's reasoning, and the evidence passages it found. Click View in context to jump to the exact passage in the patent full text and check it yourself.

If the AI got it wrong, override the coverage level with the buttons in the detail panel. You can also add your own evidence: paste a passage or link an annotation you've already made in the patent viewer.

Step 5: Export

When your analysis is complete, click Export to download the matrix as an Excel spreadsheet with all elements, references, coverage levels, evidence, and reasoning included.

→ Your first search in 5 minutes — start here if you haven't run a search yet

→ Export your results — other ways to get data out of Amplified

→ Feature Matrix reference — detailed guide to color modes, filtering, row summaries, and keyboard navigation