Oral argument — Ultratec, Inc. v. CaptionCall, LLC

The United States Court of Appeals for the Federal Circuit is now open and in session. God save the United States and this Honorable Court. Good morning. We begin this morning with three cases which have been consolidated for argument today. They are 19-1998, Ultratec v. CaptionCall, 19-2000, Ultratec v. CaptionCall, and 19-2003, Ultratec v. CaptionCall. Ms. Noel, whenever you're ready. Thank you, Your Honor, and may it please the Court. Kristen Noel on behalf of Ultratec. At issue in these appeals is the validity of eight patents that collectively cover the invention of captioned telephone in service. This was not an incremental improvement to a widget, but the invention of various captions. Ms. Noel, this is Judge Lurie. Am I correct that what you have is a limited remand on the question of whether Mr. Accio-Grosso's trial testimony and his board testimony were inconsistent? And the board decided that there was no inconsistency, and isn't that the end of your cases? Your Honor, we were on a limited remand, but that is not the end to our cases because this Court has never in the first instance addressed our appellate arguments. In other words, there's something yet to be decided, but these other issues in your briefs are really not before us, are they not? Respectfully, Your Honor, I disagree. All of the issues are properly before the Court, and the only issues that weren't in our original briefs are not outside the mandate because the mandate rule doesn't preclude issues that could not have been raised either below, before the final written decision, or as part of our first appeal. So, for example, all of the patents in this case have expired. They've been expired for three years now, but that expiration happened after this Court heard the last appeal. So, when the patents were remanded back down to the board, it was incumbent upon the board before reissuing final written decisions to construe the patents under Phillips. If the Court disagrees with us on that, certainly as this Court's precedent has been, when this Court hears an appeal from a final written decision, when the patents expire even post final written decision, claims construction must be done using the Phillips standard. Proceed. Thank you, Your Honor. Counsel, this is Judge Moore. As a housekeeping matter, before you get into the details of your claim construction arguments, can I ask you about these constitutional arguments? You've sort of basically said everything is unconstitutional, sort of throwing in every provision I can think of. And my question is, aren't most of your arguments, if not all of them, already resolved by this Court? For example, by the case of Celgene or OSI Pharma? Are these things that are properly before this panel for resolution by this panel, or are you making these arguments to preserve them in an attempt to seek in bank review? It would be the latter, Your Honor. We certainly acknowledge that the Court has issued decisions like the Celgene decision before, and you've subsequently, since our briefs have been filed, have reaffirmed that standard. And so Ultratech makes its arguments to preserve them, either for en banc review or a petition to the Supreme Court. But we do acknowledge this Court has decided the constitutional issues. Thank you. Please proceed. Thank you. Reflective of these inventions six years ago, a jury found the asserted claims valid and infringed by the petitioner. Awarding Ultratech and its affiliate CapTel $44.1 million in damages, a judgment that was later stayed. Ultratech raises a number of issues with the Board's handling of the IPRs. Time permitting, I'd like to cover two. Objective indicia of non-obviousness, and more specifically, the proof of actual nexus, the direct proof to the claimed inventions and their benefits, as that is pertinent to claims in seven of the patents. And the construction and application of trained to the voice of the call assistant, because that term is recited in six of the patents. But for all of our issues on appeal, those two included, we note that the Board did not act as a neutral, but rather advocated for invalidity, including in almost every single issue, making decisions based on arguments that the petitioner itself did not advocate below. And that practice was squarely rejected by this Court in SAS and in Magnum Oil. Turning to objective indicia, Ultratech presented strong evidence that the inventions themselves and the direct benefits that they conferred, which this Court ruled in Rambis Rhea, is appropriate to consider with secondary considerations, that they received overwhelming praise by consumers, by the industry. They also met the needs of the deaf and hard of hearing community. And they were a tremendous commercial success. We demonstrated nexus by having our technical expert review the patents, explain the claims features, and the benefits that the use of those inventions conferred. He also provided undisputed evidence of how CapTel embodied the inventions. Mr. Ludwig testified as to the tremendous commercial success that CapTel achieved and how the features and the benefits drove that success. Importantly, this testimony was corroborated by hard of hearing advocates Brenda Patat and Connie Phelps, who provided evidence of what the consumers themselves said contemporaneously about the demand, not just for CapTel generally, but for the specific features and benefits identified by Mr. Ludwig. Well, counsel, we review this question for substantial evidence. Is that correct? That's correct, Your Honor. And didn't CapTel point out that Mr. Ochio Grosso testified that it was his opinion that unclaimed features or prior art features are what drove CapTel's actual sales and that Ultratech's initial dominant market position and growth of the hard of hearing industry, hard of hearing population, drove the commercial success? So didn't the board have in front of it expert testimony that contradicted your expert in terms of whether or not it was the claimed features as opposed to unclaimed or prior art features, which actually resulted in the secondary consideration evidence that you presented? Respectfully, Your Honor, I will disagree. They didn't actually provide any affirmative evidence, Mr. Ochio Grosso. He just tried to point out what he believed Mr. Ludwig did and did not consider. And Mr. Ludwig's declaration somewhat belies that. He did look at unclaimed features and compared them to the claimed features and explained why they drove success. Caption Call also pointed to something, for example, that it entered the market and minutes of use went up. But what they failed to recognize is that a jury found that their product and service infringed all eight of these patents. So that doesn't undercut commercial success. It, in fact, bolsters it. Patat and Phelps reviewed hundreds of comments that consumers submitted to the FCC advocating that the FCC make Caption Telephone a compensable service so they could continue to use Caption Telephone. These comments were not just that Captel was life-changing, and many of them reflect that, but they identified specifically what it was about the product and service that changed their lives. The speed and accuracy of the captions, which we know is afforded by revoicing. The naturalness of the call afforded by having the ability to use their own voice on the phone, use whatever visual hearing they had, and to have the captions help keep up with the conversation. That is the method of providing Caption Telephone service with voice and revoiced text. And they also had, and complemented, the ability to have others call them directly instead of having to call a relay on a different number. And the security that they could know that they could call 9-1-1 and not have to choose between having captions so they could understand the operator or have the 9-1-1 operator lose their location. As Mr. Ludwick confirmed, this was the result of the two-line functionality with Caption Telephone. Now, we know why CapTel was commercially successful. Because the consumers themselves told us. Patat and Phelps also testified at how these specific benefits met the long-felt needs of the deaf and hard-of-hearing community. Indeed, Ms. Patat testified how CapTel was, quote, life-changing, quote, life-altering and life-saving. Counsel, one of the problems the board seems to have had, I mean, you are pointing to a lot of evidence right now. But one of the problems the board seems to have had is that it said, for example, on page 118, that your response itself contained none of these arguments. At one point, and I don't remember where in the board's opinion, but they said, you pointed to like 137 pages of declaration as though we, the board, are supposed to go on some fishing expedition looking for the kernels of evidence that may or may not substantiate your argument. And so what the board, even if there were evidence in the record that you're now pointing to, if you hadn't pointed to that evidence with particularity to the board, and if their concern was your allegations as argued to them were conclusory and you just dumped tons of evidence on them without particularity or pointing to anything specific, should we really fault them? Should we reverse them under those circumstances? The fact that you've now stepped in and are making a much better argument to us than you made to them, is that a basis for us to reverse them? Yes, Your Honor. And I do take issue with the board's characterization of the presentment of the evidence. In some of the responses, secondary considerations detail 13 pages of argument and summary of the evidence. We could not have possibly put all of the evidence of secondary considerations into the brief, but we made the argument and we cited to the declarations. And if you review the declarations, they're actually not that voluminous, but collectively there are three. And they are very well set out with headings and subheadings. They give actual evidence and detail to support the argument, which is made in the response. And the board said that it went and it considered this evidence, and its real justification for not giving it any weight was this claim of lack of embodiment. Embodiment wasn't disputed by CapTel. Let me be clear about this. Of the 37 claims across the eight patents, CaptionCall did not present, even though they had two experts, did not present evidence of a single element of any of those 37 claims that they assert was not embodied in the CapTel phone. So the board took it upon itself, acted as an advocate on an argument not made by CapTel, and that is something that violates the in re magnum precedent from this court. Who has the burden on that question? Is it either the burden of production or the burden of persuasion on that to show the connection, the nexus being drawn between the second consideration? We have the burden of showing nexus. And as this court has observed in Polaris and other cases, when we present undisputed evidence that the other side does not dispute and does not present any evidence to contradict, it is an error of the board to disregard it. Moving on to train to the voice of the call assistant. I'm sorry. Could I just get sort of a logistical question, and that is the argument you just made with respect to secondary considerations, does that apply equally to each of the three separate appeals we've got in front of us? Yes, Your Honor. And, in fact, I will note the last appeal on Claims 6 and 8 of the 835 patent, the board actually did not only deny nexus but also held that the minutes of use were not captioned telephone service because they looked at the FCC's minute data, saw the words CTS, VCO, and took it upon themselves to say that was the prior art. And so they discounted commercial success. That was an argument, once again, not made by Caption Call because it is not true. Counsel, you suggested that the board erred in response to my prior question when it said patent owner's response contains no substantive arguments. Instead, patent owner merely lists various common forms of secondary consideration evidence without exposition. Can you tell me where in the appendix I can find patent owner's response and, in particular, where in that response you believe that you have, in a detailed fashion, articulated the arguments the board should have addressed? Yes, Your Honor. For example, in the 2003 FCO, You have to cite it by appendix number for me. Yes, Your Honor, but we have three appendices. So in 2003, I would look, for example, at appendix 512, and I would also look at appendix starting at 4124, 4125, and it goes on for some 13 pages outlining the evidence. On training to the voice, Your Honor, I think our construction is fairly well laid out in our brief. I will say that, so it doesn't get lost in the briefing, the most important for the court to consider when looking at the single loan sentence in Ryan is to understand what was the state of the art of voice recognition software at the time of the invention, and specifically, I would direct the court to appendix 3131, which is from the McAllister patent, and in the 2000 case, Your Honor, Thank you, Your Honor. In the 2003 appendix, for example, I would direct the court to appendix 511 to 513, and this pertinently on page 12. I would also direct the court to appendix 4214 through 37. Thank you. Thank you. I'm sorry, Your Honor. I misread it. It's 4124 through 4137. Turning to train to the voice, which is a claim term used in six of the patents, I think the claims construction, particularly now that we're under SOAP standard, is very clear and direct. The only issue with the board... Counsel, I'm sorry to derail you, but we can't find any document with the page number 4124 to 4137 in the 03 appendix. I'm trying to actually look carefully at what you're telling me to look at, but I don't see those documents. My apologies, Your Honor. I'm looking at a stamped copy, which is the patent owner response from patent number 6603835, which is one of the patents at issue. Okay. And I will have my colleagues double-check the record. If, with leave of court, I'll move on. On train to the voice, under the Phillips standard in particular, the board's holding that train to the voice of the call system can mean train to the voice of a group is clearly erroneous. The only basis the court gave for this is a loan reputation on a line in the patent that refers to train to the voice of the voice pattern of the call assistant. Once again, this is an argument that was not presented by caption call. The board took it upon itself to make this argument. But it also holds no water. It plainly refers to a single call assistant. Mr. Otrio-Grosso himself construed this to refer to a single call assistant. Moving on to Ryan, what I don't want... Before you move on, I need to understand something. Are you suggesting that the board isn't permitted to look at the claim language or the specification and come to its own conclusion if the opposing party didn't make that specific argument? I mean, the term is in dispute. So isn't the board supposed to look to the intrinsic record to understand the meaning of the term? Certainly the board should look at the intrinsic record, but I do believe it's this court's precedence and in IPRs in particular that the board is not to take it upon itself to make arguments and hold positions that the petitioner itself has not put forward. I believe that's an in re magnum decision. Also, it's just plainly wrong. Their expert agrees. Our expert agrees. And if you read the specification in this patent, you cannot walk away with the understanding that this means trained, much less designed to a group of call assistants. The patent itself teaches that it is a limitation of the software that you have to use, voice recognition software that is trained to the voice of the particular call assistant. So the reference that is cited against us, the Ryan reference, has a loan sentence designed to recognize the voice of the particular relay agent. And I think in light of the lack of exposition in Ryan, we need to read that literally, and very importantly, we have to understand the state of the art at the time. When you look at things like the 314 patent, column 2, lines 45 to 49, it says speech recognition computer program trained to the voice pattern of the call assistant, and then it gives an EG an accent. I'm not going to suggest that all southerners speak with the exact same accent, but certainly more than one person. I'm from Baltimore, and everyone in my family says it with a D instead of a T. It's an accent. So why isn't that relevant? Your Honor, with respect, I turned to the 314 patent, and it doesn't say EG accent. It says trained to the voice pattern of the call assistant. I have a voice pattern. Your Honor has a voice pattern. Mr. Shaw has a voice pattern. We all have voice patterns. That certainly cannot, just because a group can have a shared voice pattern, and I would agree with you that a group can have a shared voice pattern, you can't trump the teaching of the specification, which is it's limited to speech recognition software like Dragon NaturallySpeaking. The word pattern doesn't mean a group. Now, very importantly, there was also at the time voice recognition software, as Your Honor described. It was not speaker-dependent, meaning you couldn't take it off the shelf and have it learn your particular voice. It was hard-coded, but it was coded to recognize the shared singular voice of a group of people, like an accent, like Your Honor just recognized. And if you look, for example, in the 2000s. Please finish your thought. Thank you. If you look at the McAllister reference found in the 2000 appendix at 3123-3137, and in particular at column 4, lines 58-67, you will see a description of existing approaches to speech recognition technology, which included universal, which is speaker-independent, voice recognition software created using speech models for samples of accents, and in particular the line, for example, if the models are created using speech samples for New Englanders, then the models will tend to exclude voices with southern accents or voices with Hispanic accents. That is voice recognition software that is designed to recognize the singular voice of multiple people. That is literally what is the line disclosure in Ryan. Thank you, Your Honor. Thank you. We'll reserve the remainder of your time for rebuttal, and it's time to hear from Mr. Shaw. May it please the Court. Prateek Shaw for APALE Caption Call. The last time these appeals were before this Court, Ultratech secured a remand in an attempt to discredit Caption Call's expert, Mr. Ocho Grosso. But that gambit failed. Ocho Grosso's expert testimony, according to Ultratech itself, provided material evidentiary support for the PTAB's unpatentability determinations. Now that the PTAB has squarely reaffirmed the consistency and credibility of that testimony, Ultratech can no longer possibly argue a lack of substantial evidence underlying the unpatentability determinations. Counsel, the prior remand, this is Judge Moore, and the prior remand, if you remind me, was claim construction at issue on appeal? In the prior, in the first appeal, yes, Your Honor. And did we actually resolve any of those questions of claim construction? No, Your Honor. You just remanded on that narrow issue. So those are, we would agree that those are before you now. So claim construction is now before us. And one thing that has happened between the time when the board issued its claim construction and now when we're reviewing it is the patents have expired. Doesn't that de facto push us into Phillips' territory? So, Your Honor. What I'm trying to say to you is forget about whether the board used Phillips or not. Sure. Forget about whether the board was proper with regard to the scope of the remand. Aren't we, our court, now clearly in Phillips' territory in light of our court's precedent? Your Honor, yes. This court's precedent does seem to indicate that it is applying Phillips on there. I think there are distinctions to be drawn here just because of the unique procedural posture of this case in that if it had been decided on the merits the first time it was here in that first opinion before the expiration, it would have been BRI, but for that kind of narrow remand on the procedural issue, which really didn't change anything because it came back up finding the testimony consistent, you're now back here in that posture. So, I think there are distinctions to be drawn just based on the unique procedural posture. It's not implicating some of the policy interests that underlie the switch from BRI to Phillips. The other side has never tried to amend even though you had these same adverse constructions now for several years. So, I think there is a basis to continue to review under BRI, but assuming this court were to apply Phillips, I'm happy to address why we think the result should not change with respect to the claim constructions that are now at issue on appeal. And if you'd like me to start there, I'm happy to start there. Well, I'd like you to – this is Judge Prost. Where I'd like you to start is a theme throughout your friend's argument, it seemed to me, was the lack of neutrality by the board in the way it handled various issues in this case, making arguments not made by petitioner. It cited our case in Ray Magnum. So, I would like you to somewhere respond to that theme and those arguments, if you wish. Sure, Your Honor. I would disagree with that characterization. We vigorously disputed virtually all of these issues before the PTAB. There's 750 pages of decisions now spanning this appeal through an initial decision, then a rehearing, and then a remand. And at each step, the arguments change slightly, and then we came back with counterarguments as the arguments evolved. There's nothing unusual about the board looking at all of the evidence and looking at all the arguments and coming to what it thought was the best decision. We made virtually all of these arguments, but to the extent the board wanted to look at the patents itself and the prior art itself and parse them on its own and not just follow the experts on both sides, of course the board is entitled to do that. The board wasn't acting as an advocate in any way. It was acting as a decision maker, just as courts do all the time. It looks at both sides' arguments, and it may have a third way of resolving the case. And I think it's helpful to do this in the context of the specific arguments. I'm happy to either start with the claim construction or the secondary considerations issue. I can start with the secondary considerations issue, since that's where the other side started, if it's helpful. And I think there's some important things to keep in mind here based upon what happened before the board. One, as Chief Judge Proz noted, it's Ultratech's burden on Nexus to show, to prove Nexus, even if we had not submitted any evidence to the contrarian, and we did, and I'll walk through that. Even if we didn't, it's still their burden to show Nexus, and the board can hold them to that burden. Now, what the board said is that they failed to show Nexus on two independent grounds. The first one is the one that, Judge Moore, you mentioned, the procedural ground. They provided, as the board said at page 118, here's what the board said. It reviewed the documents and said, Patent owner's response contains no substantive arguments, and this is on the Nexus point. Instead, patent owner merely lists various common forms of secondary consideration evidence without exposition. This does not provide sufficient analysis for us to determine whether the patent owner has provided adequate evidence of secondary considerations and a Nexus between any such evidence and the merits of the claimed invention. Thus, patent owner's broad contentions regarding secondary considerations do not demonstrate non-obviousness. Now, if you look at what it did, when you asked the other side, Ms. Noel, for the citations to their argument, she gave you the only citation from the last IPR, the ninth IPR, when they finally remedied that error and provided more than two pages. In the first eight IPRs, and the board notes this, in the first eight IPRs, it has the same language that I cited to you in the first eight. They only made two pages of cursory argument. She didn't cite you those pages. That's JA 786 to 787, and then there's corresponding cites. For the first eight IPRs, they failed to make the substantive argument. And if that was not enough, on rehearing, they tried to reissue, make the same argument in the first eight IPRs, and at appendix page A440, and this is of the 1998 appeal appendix, here's what the board said. So after the board said, look, you've only given us less than two pages, and by the way, their counsel at oral argument in front of the board said, you're right, Your Honor, we only gave you a page and a half. We should have done more. Now, on rehearing, they argued, well, look, we cited the declarations, and even though we didn't make the arguments in our brief, you can read the declarations and figure it out. Here's what the board said on rehearing in response to that at page A440, and again, this is repeated through the first eight of the nine IPRs, every patent except claims six and eight of the 835 patent. Here's what they say, and this is I'm quoting from appendix page 440. In this request, patent owner seems to suggest that we should have reviewed and analyzed the entirety of each of the three declarations submitted by patent owner in support of its secondary consideration contention. Patent owner merely cited each declaration in its entirety without citing with particularity portions of these declarations. Yes, counsel, this is where the 137-page thing comes up, right? Exactly. That's the next line. We will not scour the 137 pages of declaration evidence submitted by patent owner and generally serve as an advocate for patent owner. So this is where the board is saying we're not going to serve as an advocate for them, Chief Judge Prost, to get to your question about being an advocate. So it's not the board's job to look at a page-and-a-half boilerplate argument about secondary considerations and then read 137 pages of declarations and try to put together the argument that Ms. Noel has made quite nicely here this morning about how – Counsel, you suggested that, and I understand, and this is very helpful, it's the first eight IPRs, but you suggested in the ninth IPR they finally cured. Yes. Did the PTO treat it differently? Yes, it absolutely did, Your Honor, and that was going to be my next point. When they finally made something more than the CURS 3 argument, and I acknowledge they did in the last IPR, so that's the 835 patent claims 6 and 8 only, then in the final written decision corresponding to that IPR, and that's at the 2003 appendix, pages A224 to 238, the board gave a 14-page response, not just on the procedural forfeiture but actually on the merits, and then here's what they said on the merits. Of course, you can stop at the procedural forfeit for the first eight IPRs. For the ninth one, when you reach the merits, they said, look, we looked at the declarations you cited, and the Ludwig Declaration, which is the only declaration that goes to Nexus, what they did is they provided a claim chart to line up the patents claims with the CapTel service, and those claim charts, Your Honor, I encourage you to look at them. They're at page, for example, in the 1998 appendix, they're at page 3169. They start there. They are two-page claim charts for each of the patents. Your Honors, I'm sure, are familiar with these sort of claim charts. The PTAB is certainly familiar with these sort of claim charts. They are usually dozens, dozens of pages that go through the manuals, the pictures, the documentations, line up the experts' analysis, not just including personal observation but going through that in detail. Here you have a conclusory statement in each one that says, I observed these matters here. The PTAB said that's not enough, and in fact, Your Honor, we provide a site in our brief. You can contrast that claim chart. Once they got these adverse PTAB decisions and these further Ultratech appeals that are not before you, what we call Ultratech 4, we provided a site there. They did a claim chart. It's 190 pages. These are two-page claim charts, conclusory. And the reason why, Your Honor, that matters here is because CapTel is not a monolithic service. It's not just one phone or one product over this 10 years. There were modifications. There were different models used. And so it's not enough to simply say based on your observation during a visit that happened at the time of this litigation to say that this has always encompassed all of the claims, at least not without providing documentation, going through the manuals, going through each and every model. They don't do it model by model. They treat it as a monolithic. And the PTAB, and this is at 2003, appendix 229, page 229 in the 835 appeal where they finally made the arguments and not just forfeited it, there the PTAB calls them out on this, and they say, look, this is not a monolithic service. And, in fact, the Model 100, the first phone that was at issue here, did not have two-line service. So there were three features that Ultratech pointed to as the breakthrough features, the two-line architecture, the revoicing, and the simultaneous delivery of voice and text. So two-line architecture wasn't there. So it wasn't, in fact, co-extensive. It was both under-inclusive and over-inclusive. And, Judge Moore, this goes to your point. There was, in fact, expert testimony by our expert, not the PTAB, but our expert submitted a declaration, which is at pages A2569 to page A2576 of the 1998 appendix. Hey, counsel, I appreciate your very clearly amazing, detailed understanding of every single page of nine appendices. How about you move on to claim construction, because my eyes are glazed over with all these page sites. Okay. Sorry, Your Honor. Okay. To claim construction. So on Ryan, Your Honor, I think let me start there with that claim construction. I'm happy to walk through the three claim construction issues. On Ryan, I think it's helpful just to take a step back. Ultratech's actual claimed invention is not the software itself. It's not claiming an invention over voice recognition software. Its invention is on the relay method of using a call agent to do real-time revoicing using commercially available speaker-dependent software. That is exactly what Ryan discloses. Using, in the words of Ryan, an agent to listen to the caller and repeat the voice message using, quote, software specifically designed to recognize the voice of particular relay agents. Now, as Ultratech concedes, by the time of its patent, the speaker-dependent speech recognition software that allowed one-to-one tailoring was already there on the market. And so this is not an invention, even though they like to try to claim it as software that only does this after the fact training. The invention is about the real-time revoicing here, and that's precisely what Ryan discloses. Now, Your Honor, on the particular claims, they talk about the one agent. I think we've already made arguments in our brief as to why the PTAB did not air, and I don't think they aired even if you apply Phillips. But even if you have doubt, probably the easiest way for this court to resolve this issue is say, fine. Even if you accept Ultratech's claim construction, that their patent, despite the language that you point to, Judge Moore, about talking about a voice pattern and an accent, even if you accept arguendo that their claim was limited to one specific agent, at page A95 – sorry to give you another site, but here's what the board does – at page A95 of the 1998 appendix, and this is the final written decision of the board that I'm quoting from, here's what it says. It addressed the construction in the alternative, and here's what it found. Moreover, we are not persuaded by patent owner that a person of ordinary skill in the art would interpret Ryan as only disclosing software written specifically for a group of people. So you have a finding by the board at page A95 that even under Ultratech's claim construction, Ryan discloses it. And there is ample expert testimony. This is not the board acting as an advocate, Chief Judge Prost. Again, I'll give you the JA sites because that's where our expert testimony is. Specifically on this point, JA – and this is the first 1998 appendix – 3557 – these sites, by the way, are all in our brief. But JA 3557, Ocho Grosso says, when Ryan says particular relay agent, that means, quote, one specific agent. Again, there's another site, JA 2554 in the 1998 appendix, where again, substantial evidence. The expert says, yes, this refers Ryan discloses voice recognition trained to one call assistant. So the board made the finding at A95 that even under Ultratech's claim construction, Ryan discloses it. There is ample expert evidence to support that. Our expert says exactly that. Now, once you accept that, if you accept that it means one agent, then Ultratech itself argues. This is their own reply brief at page 12 of their reply brief, and I'll just read you what they wrote at page 12. Indeed, CaptionCall does not contest that if this court finds that the software must be trained to individual voices – and that's the argument that I'm saying. I'm happy to accept for purposes of this argument that it's an individual voice. Here's what Ultratech says. Quote, the software must likewise be trained after development. So what Ultratech says is a logical matter. If you agree that either the patent or Ryan discloses that it reads on just one agent, then as a logical matter, of course the training has to happen after the fact. And so Ryan necessarily discloses because it reads on one agent. It necessarily has to disclose the sort of training that Ultratech says is claimed. So again, you can accept Ultratech's claim construction that the training, even though the patent doesn't say it, the training has to happen after the fact. Even if you accept that fact, well then again, we have expert testimony. I'm sorry, Your Honor. Just finish your sentence because the bell has rung. Oh, okay. I'm sorry. I'll just give you the sites so you can see them. It's at 1998 JA 2554, 3553 to 3554. Those are the sites from Ultratech's – I'm sorry, our expert saying how training happens after the fact because this Ryan discloses speaker-dependent software that has an algorithm that allows for training after the fact. Thank you, Your Honor. Ms. Lynch, were you here to just argue specific issues with respect to the constitutional? Other questions? Yes, Your Honor. The scope of the remand as well. Okay. Proceed. This court's – may it please the court – this court's remand order was narrowly tailored. First, the board was to determine whether Mr. Okio Grosso's testimony was inconsistent. And if it was, the board was then to consider the impact of that testimony. Counsel, even if we agreed that the remand was limited, don't – what is the PCO's position on what our case law demands we apply in this current appeal as to the standard of review? So forget about whether the agency erred by not reviewing claim construction. This is our first time reviewing claim construction and the patents have expired. What does our precedent say the standard is we ought to apply? So, Your Honor, in the Apple v. Andrea case, this court did use a Phillips claim construction when the patent expired on appeal. But in that case, unlike here, the parties had agreed the claim construction would be the same under Phillips or BRI. And so, respectfully, as a reviewing court, we believe that this court should review the board's BRI claim construction to see if it's correct. And you think that's true any time a patent expires between when the board renders its decision and when this court hears the appeal? You believe that we should still be reviewing under BRI even though the patent expired? Well, yes, Your Honor. And obviously, if you review the board's BRI and you find it's incorrect, then we think the court should apply Phillips. But in the first instance, as a reviewing court, we think you should review the board's claim construction. I don't understand what your answer just was. So you're saying if we look at BRI and we say they shouldn't apply BRI or we just say that the construction was erroneous and then they get a second bite under Phillips? I'm not clear on what you're saying. Right. The second, Your Honor. So as a reviewing court, we think if the board used BRI and it correctly used BRI because the patent hadn't expired at that time, we think that you should review the board's BRI construction to determine if it's correct. If you find that the board's BRI construction is erroneous, then going forward, this court, because the patent has now expired, would use a Phillips construction. So, Ms. Lynch, I've got to be honest. I thought that this was pretty well settled, and I thought that it's understood that if the patents expire at any time during the process, everybody flips over to the Phillips standard. Even in our CFB case, the board itself recognized that, and the board had in front of it a patent which had been reviewed under BRI, but then the board, on rehearing, turned around and redid it under Phillips, recognizing that if the patent expires at any time in the process, people should pivot towards Phillips. I guess I've got to be honest. I thought this was such a well-settled point, but if the PTO really believes that we have to review BRI even when the patent expires in between, maybe there's a reason for a precedential decision in this case, because that just seems to me completely inconsistent with our approach consistently applied through a number of cases on this general point. What are your thoughts about that? So, we agree, Your Honor, that at the agency, when the patent expires before it gets to the board, before a board final written decision or during reexamination, the agency will apply BRI. But like we said, as a reviewing court, if at the time the agency applied BRI, it was the correct standard to apply, then we think in the first instance this court should review to see if that BRI analysis was correct or not. In other words, we're reviewing the decision that was rendered. Correct. Correct, Your Honor. Can I just ask you a practical question? Sure. I don't recall the dates. You may have them offhand when the PTO switched course and went to the Phillips standard, so I'm wondering how much is left in the pipeline that would implicate this question. Do you have any sense of that? I know it's been a while, Your Honor, so I don't know how much is in the pipeline, but I agree with you. There's less and less as time goes on. But Ms. Flint, I'm just troubled because when you look at the Apple versus Andrea case, we didn't caveat it when we entered the following holding. This is the holding. When this court reviews the claim construction of a patent claim term in an IPR appeal after the patent has expired, such as in this case, we apply the standard established in Phillips, not the broadest reasonable construction. It doesn't seem to me that we put any caveats at all on that clear, broad statement in Apple. I mean, I'm honestly kind of surprised. I'm not surprised that you argued to me that your remand was limited, but I'm genuinely surprised that you think it's still somehow an open question as to what standard we ought to be applying right now. So, Your Honor, I agree. That is the language of Apple versus Andrea, but, you know, there is another precedential case. It's the Celgene case, and the patent had expired before the decision. That got brought up in the briefing. The court didn't address it in its decision that the patent had expired, but in that case the court did review the board's BRI construction de novo. Can I ask – I'm sorry to prolong this, but I just have to go back to one point you made, which I was just confused by. In a normal IPR, if we disagree with the board's BRI construction, is it the office's view that we then have to alternatively say, well, it would have held up over Phillips? Or are you saying that because of the posture in this case, even though you don't agree with us, as Judge Moore said, that you should be applying Phillips, that there's something different about this, so you get two bites at the apple in this circumstance and not one? I'm sorry, Your Honor. I'm not sure I completely understand that question. Could you repeat it? Okay. Actually, given the interest of time, I will just drop it and conclude your argument there. Thank you. Ms. Noel, you have some rebuttal time. Thank you, Your Honor. A couple quick points. Responding to Mr. Shaw, I did give you two citations from the patent owner responses. One was from one of the first eight, and the last was from the ninth. So I take issue with his characterization of what I represented to the court. Secondly, all this discussion about Mr. Shaw. The first one you gave us was just really page 512, and I agree with the agency that that one, which is the one that was not in the last one, is completely conclusory. It amounts to literally one paragraph in which you say nothing other than objective consideration, support non-obviousness, including commercial success, failure of others, long-felt need. There's no particularity of any kind of argument on page 512. That was why I was pushing hard to look at the other pages, which I still don't seem to be able to locate in the appendix that you say they're in, but because this page 512, even if it appeared in all of the first eight, is completely conclusory, and I could not fault the PTO for its position. Your Honor, if I may address that, we view this as a roadmap to the declarations. When you look at the declarations, they all have headings and subheadings, and they very clearly articulate the evidence here. Mr. Shaw seemed to indicate that we cured in the ninth IPR, and that the board then dove into the evidence. The board came up with the same conclusions of the ninth, 13 pages. Still wasn't enough for them because it wasn't the arguments in the response so much as they didn't like the conclusory nature of Mr. Ludwig's testimony and his claim chart. All of that, by the way, Your Honor, goes to embodiment. Embodiment is undisputed in this case. They had two experts inspect everything about CapTel. They haven't come up with a single limitation, a single element that is not embodied. Embodiment is not contested, and as this case, this court held in Polaris, observations by Mr. Ludwig, you know, is there a headset? Do captions appear on the screen? I don't need a manual for that, and you can't have a manual for, you know, the whole system. He observed how it worked. He worked in the industry for decades, and they don't point to anything he was wrong on. Turning to Ryan, there's a fundamental disagreement here. When you are talking about voice recognition software that is trained, it has to be to an individual, and Mr. Otrio Grosso agrees with that. We agree with that. The patent owners taught that in the specifications. When you are talking about designing software, you can design to the shared voice or dialect of a group of people. That is what is in McAllister, and that's what's disclosed in Ryan. Thank you. Thank you. We thank both sides, and the cases are submitted.