For me, one of the sparks for building an AI-based patent search engine was the frustration when relevant prior art popped up too late in the patenting process.
It does not feel great if a highly relevant patent – that was there all the time and not even hiding – is brought up by a Japanese patent examiner after four years of prosecution and countless hours of work. One cannot avoid the question: “What if I (or my client) knew about this patent 4 years ago?”
I know that sometimes filing a patent application has marketing value in itself, even with little or no chances of success. And you can conclude that at least the disclosure became prior art for others. But these are exceptions. Usually, companies want enforceable patents that give them a competitive advantage in the long term. And money should be spent wisely, not on impossible cases.
Back in the day, there were not many alternatives. Extensive patent searching with classical methods for every invention was out of the question, both for the attorneys and their clients. Simply because no one had the time or budget for that.
But things have changed for the better. AI enables this use case and also unlocks or changes a bunch of others that add a lot of value for organizations.
“Do work that is meaningful”, was the slogan of the Finnish army at the time I worked as a patent attorney. I sometimes wondered how this was reflected in my work.
So much wasted work or extra work to overcome the objections of the patent examiner. Most of it could have been avoided with a better understanding of the prior art. Instead, I was “drafting in the dark”.
Solution:
Searching for prior art at the drafting stage with IPRally’s claim-based search is so easy as fast that there are no excuses for not doing it. The system has been designed to ingest claims and predict the most likely citations of the patent examiners.
Moreover, the review assistant Ask AI helps you extract detailed information from the search results. It allows designing the claims around the prior art and saves you from including embodiments and variations to the draft that don’t add any value in the end.
As a result, your drafts will be more to the point and more likely to proceed smoothly. And if you’re the attorney, you’ll get happier clients.
During my attorney days, submitting third-party observations to patent offices was possible, but the option was rarely used. This was in spite they are – in principle – a very efficient tool for limiting competitors' scope of IP and ensuring your own freedom of operations.
The reason was that submitting an observation requires A) prior art and B) reasoning.
Solution:
A smart AI patent search provides A) prior art and B) reasoning. Oh, wait! Wasn’t that exactly what we needed?
Claims in, relevant results out.
Questions about key features in, human-level reasoning for novelty and inventive step out.
Third-party observation in. Best case, no patent out.
Common IP risk management processes in corporations include regular monitoring of competitors’ published applications and granted patents. In addition, there are monthly or quarterly IP reviews where it is decided if any of the patents found form a threat to the business. If so, it needs to be decided what to do: live with it, design around it, get a license, or perhaps oppose the patent.
Both monitoring and evaluating the strength of the patents has been tricky and many organizations are facing capacity issues. This has forced them to omit some parts of the market, focus only on the biggest competitors, or do very superficial reviews – and consequently make uninformed decisions.
Solution:
AI offers a new way for:
Therefore, if desired, any company could pick are more proactive IP risk management strategy and be well-equipped to implement it in practice.
You would imagine that in large companies, in particular those that are “IP heavy”, the patent records are always accurate and tidy.
The reality is that the global patent data is increasing much faster than the companies’ resources. Most organizations are struggling with managing both their own portfolio records and competitor portfolio records.
Also, a classification system or other metadata that once made sense is not making sense anymore. And the people who once understood some area of technology well may not be there anymore.
Imagine being able to derive any information from patents – either individually or from larger data sets – by just asking in normal language.
Solution:
Let’s do exactly that: just ask using normal language!
IPRally’s Ask AI feature and its big brother Multi-patent Ask AI can ingest huge amounts of data in seconds to answer any question you may have and output in a digestible format. Tirelessly, patent by patent, question by question.
No need to re-read vast amounts of boring texts anymore. Answers are what we need.
And when a new need arises, just do it again.
It’s not uncommon that IP people groan with a twinkle in their eye about the projects they are working on and workloads waiting around the corner. Both with a genuine satisfaction about being part of something meaningful, but also to express that it would not hurt if some tedious tasks were a bit more efficient and fun.
Solution:
Test new ways of working. Test new tools. Use AI!
With efficient searching, review, monitoring, and classification, there is not only a higher probability of getting the daily success quota filled. You may also find yourself making a positive change in long-term productivity and happiness at work!
In this inaugural episode of RallyCast, myself and John Paul Keeler bring up the most important use cases for discussion and share their insights on topics such as consolidated drafting and searching, third-party observations and proactive risk management. You don't want to miss it – watch it now!
Copyright © 2022
— IPRally Technologies Oy.