AMBERCITE - AI                                                  

Ambercite Ai uses unique algorithms to find the most similar patents to one or more starting patents. Results are ranked by similarity, and searches are very quick and easy thanks to a simple and intuitive interface.

How similar? We (and our clients) have benchmarked our algorithms against the best conventional and semantic search engines we could find - and strongly believe that we have the best similarity engine out there. Our ranking is assisted by our unique family to family analysis - and is not weakened by the limitations of inconsistent language and terminology inherent in other approaches. 

What does this mean to our users? Fast, cost-effective and intuitive searches that will instantly find you similar patents missed by others.

And now with a smart review interface that allows you to very efficiently review representative images and abstracts.

Check out the demonstration video below:


Example of Ambercite Ai in use - and its unique ability to ZOOM INTO UNCITED prior art that mattered in litigation

Consider the case of EASTMAN KODAK CO., ET AL, vs CTP INNOVATIONS, LLC,, IPR2014-00788, published on 13th January 2017. In this case, a number of claims  of US6738155 were invalided in the light of Dorfman, WO199808176, and some non-patent literature. The patent claimed remote printing of PDF images

Running the prior art query shown below: on US6738155 returned 250 prior art documents.

US6738155_Ambercite query.gif

While all were potentially relevant, we can then apply keyword filtering to reduced this data set. A key feature of the US6738155 patent was the printing of PDF documents. For this reason a keyword filter was applied to the result set so that only prior art patent families that referred to "PDF" in either the title or abstract were retained. 

PDF word filter.gif

By limiting the results in this way, only one patent was left to review - being WO199808176, the knockout Dorfman patent. Yet this patent had not been previously cited against US6738155

This, in two simple steps, Ambercite was able to uncover the previously uncited prior art patent that helped to invalidate US6738155. This truly is an AI approach to patent searching - that works.


Because Ambercite Ai is so simple and easy to use, and can produce such unique results, our users increasingly agree: 'No search is complete without it'. 

Prior Art Searching

Start with a patent number or a list of known prior art patents. Get a ranked list of highly similar patents - including similar patents missed by others. Check out a video of an invalidity search below. 


Find very similar patents and their owners, for a patent or group of patents you wish to commercialise. Use our"Licensing Potential" metric to prioritize opportunities and patent owners. Check out a demonstration video of a licensing search a bit further below.

Instant patent landscaping

Start with even a small list of the known patents in the area, and quickly build a landscape from there

Portfolio Benchmarking

Separate the gems from the junk. Rank and understand a portfolio of patents.


AmberScope combines 150+ million patent citations to show you the most similar and most important patents connected to your starting patent. A variety of features can help make the results shown to you in a fully interactive webpage, as shown in the video below: 




Want to know how to interpret the bubble chart? Only 5 things you need to know. Read more below:

  • The most similar patents can easily be recognised as those with thicker lines connecting them to the focus patent (starting patent).
  • Important patents in the same area can easily be recognised as those with the largest bubble sizes.
  • Simple filters let you focus on patents that are:
  1. within a target date range
  2. the most similar to the focus patent
  3. the most important patents connected to the focus patent
  4. or a combination of the above
  • Potential infringements can easily be recognised with arrows showing the ‘flow of ideas’ through the patent network.
  • "Ghost patents" shown opaque, are indirectly linked patents, that our algorithms think are relevant to your search, but were missed by the examiner or applicant. 

Want to hear what users are saying about Ambercite? Check out our Testimonials.


For an overview of Ambercite, check out: - First Steps

There are a number of products and firms out there that also offer patent searching and analytics.

So why should you choose Ambercite products or services?


We identify what are likely to be the most similar and most important patents close to the patents that you are interested in, using advanced methods based on our AI algorithms. This saves you time and money otherwise lost on reading low value patent documents.  

Abstract, representative image and smart review features can then allow you to hone into the patents that you are looking for. This reduces the risk of missing key patents.


We provide intuitive and, in the case of AmberScope, interactive graphics. This helps make patent searching more interesting! Visuals, rather than log lists of data, are more intuitive to many people and can improve your patent searching productivity. It is thought that our brains can process visuals 60,000 times faster than text.


All our products rely on objective and automated algorithms analyzing what can be hundreds of thousands of data points for one similarity assessment. Even better, our AI algorithms analyse over 150 millions independent examiner or applicant decisions, removing the need for the risky assumptions of classification systems or subject matter text analysis. Better and more comprehensive data means better outcomes.


By focusing on being the best citation based patent insight provider, our products are designed to complement your existing search processes, and so provide a second and what may be unique opinion.