At our UNSCRIPTED roundtable event in London last year, one of the most fruitful and lively debates concerned the potential for data and technology to change the IP profession. Over the past year at Adamson & Partners, we have started to incorporate patent analytics into our own projects. Attorneys may use such tools to monitor competitors and inform their appeals strategies. But in our case, we use them to identify and target attorneys working in the precise technical fields of interest to our clients.
Recently, Adamson & Partners sat down with Tony Afram, founder and CEO of ipQuants, to learn more about the value that patent analytics can bring to the work of patent attorneys, and to see how he envisions technology continuing to influence the profession in the future. A European Patent Attorney himself, Tony began his career training at Grünecker, before most recently serving as Chief IP Counsel and Chairman of the Board at Leica Microsystems. After endeavouring throughout his career to deploy data when he could in his daily decision-making, he believed a gap existed in the analytics market which he decided to fill.
JW: Tony, thank you for your time today. Over a year ago now, you left Leica to devote yourself full time to a project I know you had been developing for a while. It is extremely rare to see patent attorneys take the entrepreneurial step of founding their own company. Tell us how you got here. Did you expect to arrive at this ambition in your career after becoming a chief IP counsel?
TA: When I was growing up, my family owned and managed their own restaurant in Sweden. By going to university and pursuing the career I did, I was doing something very unusual to them. But I guess that meant I always had the idea of starting my own company at some point. I always wondered what it would be like to be entrepreneurial. The only reason it took so long is that I was fortunate to work for so many great companies, with great people on fascinating projects. But eventually, my passion for data analytics and building this product inspired me to take the leap and leave Leica. The way I see it, you can only develop something great by also risking failure. Some of my friends thought it was crazy to take this leap at this stage of my career, but to me it was really the right thing to do.
JW: Tell us more specifically about ipQuants. Data analytics in the legal industry is still young but not new by now. Several established service providers already exist. How did you use data in your own career as an IP counsel, and what gaps did you see in existing products?
TA: I have always been fascinated by data analytics. My major was in optimisation theory at KTH in Stockholm. My master’s thesis involved optimising supply chains at Daimler, which was a pure mathematical problem and involved a lot of coding that I really enjoyed. When I started at Grünecker, sometimes I tried to find out if anyone in the firm had evidence on how certain arguments were received by examiners in the past, before I wrote my office actions. But obviously it’s hard to dig too deep when you’re training and managing workload.
But when I moved in-house, suddenly I was surrounded daily by people in R&D making data-driven decisions, and they wanted that from their IP department too. So I always incorporated data analytics into my advice as much as I could to make it credible. And when you identified the right data and gave it to stakeholders, it was amazing how they used it in creative and unforeseen ways. It turns out there are all kinds of information hiding in patents with productive purposes. For instance, during one of my first in-house jobs, a sales engineer and myself worked out who the right decision-makers were at a potential customer, based on the inventors listed on their patent filings. Another time, I conducted due diligence on the portfolio of an acquisition target, and realised lots of patents had been filed but few were granted. But there were all kinds of reasons why that could have been the case. What was clear and mattered much more was that they had a prolific inventor who was regularly cited. You could infer from the patents that the R&D talent had real value.
But the problem was that the tools out there for these kinds of analyses were often incredibly time-consuming. They didn’t allow you to zone in easily on the things you need from a more strategic point of view. What I was really interested in were the strategic insights I could gather from the data.
JW: So the main barrier to the use of existing products was that they were time-consuming?
TA: Well, I guess that was a large part of it. You don’t have time to spend weeks cleaning data to make it useful. CTOs expect answers from their chief IP counsels much faster than that. And they might not need or want an answer with a hundred percent certainty. If they did, they would pay an outside firm for a thorough analysis. They are comfortable with uncertainty so long as they know what the level of uncertainty is. But they still need more informed insights from their internal counsels than it is often easy to give without good data, and they need it quick whilst it’s still relevant.
But on reflection, the problem isn’t just time. Sometimes it’s just that the data doesn’t exist in a way that is accessible even if you spent lots of time on it. So part of what we have done at ipQuants is clean and create useful data which can now be used by our clients. When I was a chief IP counsel, I didn’t have that resource. If a CEO wants to know what the probability is of winning an appeal or a certain case, all you have is your gut feeling. But there are lots of reasons for thinking gut feelings aren’t reliable.
JW: So much of the professional services industry – not just IP law firms, but all law firms, and then also strategy consultants, investment managers and perhaps even headhunters! – rely on gut judgment. A lot of people still believe human intelligence is not reducible to hard evidence.
I know that your latest product, providing insights into appeals processes, including the precise case law that each board member often cites, is unique to ipQuants and required heavy-lifting in cleaning and formalising the data to feed your software. How do you envision this changing the way attorneys work?
TA: We started off focusing on the appeal stage at the EPO because that was an area I thought insights were lacking in and we could really drive better decision-making. It’s also an extremely important part of the patent process. This is the last chance to save the patent. You may have spent ten years prosecuting the patent, and more importantly, there’s all the original R&D spend that gave rise to the patent in the first place. So this is a make or break moment. You don’t want to go in blind. Imagine walking into a vital business negotiation without researching who you were meeting!
We have built a tool to allow attorneys to really learn about the appeal board and precise examiner they are facing. We have created transparency around whether a board faced a similar situation in the past. If they already considered an argument you intended on making and it failed, you can avoid making the same mistake again. Which benefits both you and the appeals board, right? Because you are not repeating arguments that they don’t accept. Attorneys and examiners share responsibility for making the process more efficient, and this should empower attorneys to make better arguments. In the same way medicine is becoming personalised, I believe the legal market must adapt to personalised services too.
We’re about to launch another product which I think is really interesting. For instance, if you look at our data to see which applicants are most often receiving article 56 objections, they usually receive them from a certain set of examiners. But then if you analyse article 84, a different set of examiners are prominent. We believe these kinds of tools can make processes more efficient and transparent, which will secure better outcomes for our users and the patent office.
JW: Without divulging too much, give us a sense of what the future holds for ipQuants. What other products do you expect to roll out? What has the client feedback been so far?
TA: We are going to continue to quantify and digitise all of the European Patent Office’s data, and we will soon be launching a new product that’s laser-focused on strategic insights. We’re preparing something else that I’m particularly excited about over the next few years, but I won’t talk publicly about that just yet. But overall, what really excites us is our plan to quantify everything, and when we can then connect all the results of that in a single platform, I think that will be very valuable and doesn’t really exist today.
Client feedback has been incredibly useful. Some of the feedback is that our product will completely change the way they work. They didn’t know initially if this was good or bad! But they were surprised by how much impact our tools could have on decision-making. That was good to know – that we were doing something distinctive that gets people’s attention. But obviously we also want to make sure we are useful. Fortunately, when we give people demos and they spend time playing with our tools themselves, typically they do quickly conclude that they want permanent access to our tools. So that has been great to see; that once we can show attorneys what our tools can do, they want to get their hands on them.
JW: Thinking more broadly beyond data analytics and turning to the potential of artificial intelligence, I’m curious which parts of a patent attorney’s role you see as most ripe for AI-assistance. It seems to me that there is still significant potential to automate clearance and freedom-to-operate work, in the same way discovery in litigation is now driven by search algorithms. On the other hand, I assume that insofar as they involve advocacy and oral proceedings, oppositions are safe. But what about activities like patent drafting? Do you see technology contributing there any time soon? When AI will soon drive cars, diagnose diseases and predict protein structures, it seems inevitable.
TA: Yes, I am sure it will be possible to automate the patent drafting process, at least partially. The skeleton of a patent clearly involves some repeatable tasks that could be pre-populated, before an attorney comes in and fills in the gaps.
It is hard to predict timelines. Even if the technology was there today already, there could be other barriers to accepting and incorporating these tools. But in my view, there is so much more low hanging fruit to grab first. Much simpler technology and analytics could improve how attorneys could work already today. And maybe this would be a stepping stone to accepting the more advanced technology later on.
Let’s not forget, though, that patent drafting is one of the core tasks of a patent attorney! Drafting a great patent application is not simply about taking the invention disclosure that R&D send you and starting to draft an application. Attorneys must read between the lines and align the invention with what will provide competitive advantage. This usually requires several rounds of discussions with at least the inventor, and sometimes also the product manager and marketing colleagues, after reviewing the invention disclosure. It will take a lot for artificial intelligence to grasp that kind of nuance and learn those subtle considerations any time soon.
JW: Thank you again for your time.
TA: Thank you!