Our Product Evolution

When sharing a search case or collection, it is possible to leave a message for the recipients. This allows you to provide context regarding why you share the documents (for example in relation to a specific search project or use case), enabling the recipient to take more informed action on the request.

When you leave a message, your colleague can see it in the notification menu alongside the shared content and in the email notification.

We made the Boolean search experience better. Now, the search is faster and we store the results so that next time you open the search case, the results are unchanged and load instantly. Naturally, you’ll also have the option to refresh the search for the most up to date results.

Searches without wildcards are capped to 50 000 patent families and searches with wildcards are capped to 5000 families, which should satisfy most common needs. Note that full Boolean filtering of AI searches is also available.

The stand-alone Boolean search is still under a Beta flag until we have the Boolean monitoring implementation in place too, and have polished a few things. We are open for feedback.

You can find more information on the Boolean capabilities in the Help Center.

Our unique Image search feature allows you to start a patent search with a photo, technical drawing or flow chart, for example. The feature provides a quick start for e.g. novelty, state of the art, FTO or infringement searches, when the image is the best or only starting material available. Combining Generative AI and our core Graph AI in a powerful way, Image search finds similar technology quickly and even from patents without any images. Read more in our Blog.

Now, you can easily access the information of any family member using Ask AI, just by clicking the Ask AI icon next to the family member. You can both start a new conversation or switch to an existing conversation.

Our new AI model, called Graph Transformer 3.0, understands patents and technology again better, and has been released for all users and all search use cases. The model improves search recall by about 10% and cuts the review time needed by even more. Also the custom AI classifier takes advantage of the new model. The model, like our previous models, has been trained using Examiner citations across all fields of technology, and patents converted into knowledge graphs.

Now your work with creating perfect highlights is not wasted, but you can re-use them over multiple cases. Creating new presets and editing of existing presets, for example to add additional keywords, has been made easy.

Multi-patent Ask AI extends the capabilities of Ask AI to large data sets: search results and entire Collections. You can ask the same question from all of the documents in a list and get answers in seconds. The feature allows for conveniently reviewing the search results in deep and detailed level to narrow down to the most relevant hits. In addition, the feature enables for example summarization of search and monitoring results for delivery to R&D and (re-)analyzing complete patent portfolios for strategic decision-making.

When leaving a comment, users can choose whether it is private or public. You can now change the default preference from Settings—Comment settings.

Ask AI allows for asking questions about individual patents and get clear and concise answers using Generative AI model. The feature enables a variety of use cases from customized summarization to technical feature extraction. We have paid attention to keep the answers factual and minimize the probability for hallucinations. The feature is available for all search results and documents saved to Collections.

Help Center covers IPRally's main releases since 2018, therefore going beyond this page.

See also other useful articles in the Help Center!