[00:00:00] Speaker 02: Our next case is applications in internet time versus sales force, 2024-1133. [00:00:29] Speaker 02: Mr. DiVincenzo. [00:00:32] Speaker 02: Good morning. [00:00:33] Speaker 01: Good morning, Your Honor. [00:00:37] Speaker 01: May it please the Court. [00:00:39] Speaker 01: The district court's findings of non-infringement and infidelity on summary judgment were wrong. [00:00:46] Speaker 01: With respect to both of those findings, the district court simply decided the factual issues in dispute and failed to give appellant all reasonable inferences. [00:00:57] Speaker 01: I'm going to begin with claim construction. [00:01:00] Speaker 01: We start with the language of the claims. [00:01:02] Speaker 01: That's where we're always supposed to start. [00:01:04] Speaker 01: The district court's opinion does not address the language of the claims. [00:01:09] Speaker 01: The language being construed says, automatically detecting changes that affect an application. [00:01:17] Speaker 01: It is not limited to a particular structure on its face. [00:01:23] Speaker 01: There are also dependent claims. [00:01:25] Speaker 01: The dependent claims that depend from that add one single limitation. [00:01:30] Speaker 01: wherein the automatic detection occurs using a special program, an intelligent agent that exists in the dependent claims. [00:01:39] Speaker 01: So the claims on their face suggest that they're not limited to intelligent agents, both under the ordinary meaning and based on claim differentiation. [00:01:49] Speaker 01: Now we turn to the specification. [00:01:52] Speaker 01: Appellees argue that the problems to every aspect of the invention in the patent [00:02:00] Speaker 01: limits all the claims to intelligent agents. [00:02:04] Speaker 01: That's not remotely the case. [00:02:07] Speaker 01: The abstract is a paragraph long. [00:02:09] Speaker 01: It talks about finding changes. [00:02:11] Speaker 01: It doesn't mention intelligent agents. [00:02:15] Speaker 01: The summary of the invention doesn't mention intelligent agents. [00:02:20] Speaker 01: And this court has found a limitation that's seeking to be read into the claim. [00:02:26] Speaker 01: And that's what they're doing, because they're not construing the ordinary meaning. [00:02:29] Speaker 01: You're just reading it in. [00:02:30] Speaker 01: It's hard to say the invention as a whole requires that limitation when it's not even mentioned in the summary of the invention. [00:02:39] Speaker 03: There's one sentence, I think. [00:02:42] Speaker 03: where I don't know which column it's in. [00:02:45] Speaker 03: But both the district court, if I remember correctly, and your adversary rely on it. [00:02:49] Speaker 03: It says the invention, something like that, with respect to the intelligent agents. [00:02:55] Speaker 03: That's probably the strongest evidence. [00:02:57] Speaker 03: But what is your response to it? [00:02:59] Speaker 01: That is in column 10. [00:03:01] Speaker 01: And it begins at column 10, line 26, is where the language begins. [00:03:08] Speaker 01: And it says. [00:03:10] Speaker 01: The following example illustrates how a change made to a regulation is identified on the internet [00:03:19] Speaker 01: and incorporated and managed by the invention. [00:03:23] Speaker 01: Now, that's using the word invention. [00:03:25] Speaker 01: But this court, when it looks to the specification and uses the invention, it's not all embodiments require the inventions required to contain this. [00:03:37] Speaker 01: It literally says the following example illustrates. [00:03:40] Speaker 03: So that's why the sentence I was thinking of was in column 10, starting line 40. [00:03:45] Speaker 01: Yeah. [00:03:47] Speaker 01: And then it says, a example. [00:03:49] Speaker 01: Then there's a paragraph setting up the illustrative example. [00:03:52] Speaker 01: And then the next sentence says, the invention begins tracking changes using one or more intelligent agents. [00:04:00] Speaker 02: But is there anything in the specification using other than intelligent agents? [00:04:06] Speaker 01: No, intelligent agents are the best mode of the invention. [00:04:11] Speaker 01: But both the summary of the invention and the abstract describe collecting changes without using the term. [00:04:18] Speaker 01: So that's a particular feature, as you would expect in the dependent claims. [00:04:23] Speaker 01: But they're trying to read it into all the claims. [00:04:26] Speaker 03: Can I ask you something else? [00:04:27] Speaker 03: This word, intelligent agent, is defined in Columns 10 and Columns 20. [00:04:33] Speaker 03: And for example, Column 20 says, [00:04:39] Speaker 03: based on predefined rules and objectives. [00:04:42] Speaker 03: So I'm wondering, that's a really broad, and I'm wondering why does this matter? [00:04:49] Speaker 01: It matters because the judge granted summary judgment of infringement against us on the merits. [00:04:55] Speaker 01: I didn't think it mattered at the claim construction hearing. [00:04:58] Speaker 01: We didn't think it mattered until we got the summary judgment order. [00:05:01] Speaker 01: If we look at the evidence, on the evidence itself, they conceded below it's a specialized program. [00:05:09] Speaker 01: They put forth no contrary evidence that it has rules, objectives, tasks, constraints. [00:05:17] Speaker 01: There's just no other evidence. [00:05:18] Speaker 01: And our expert went through and summarized the tasks and objectives and concluded it meets all these definitions. [00:05:25] Speaker 01: Now, it does have a meaning in the art, though. [00:05:28] Speaker 01: And the meaning is consistent with those. [00:05:31] Speaker 01: And the language actually distinguishes it. [00:05:34] Speaker 01: Because, OK, we're talking about something with tasks and objectives with respect to automatically detecting changes. [00:05:39] Speaker 01: And that's what makes it very different from POP. [00:05:43] Speaker 01: And that's why Salesforce's expert didn't apply the ordinary meaning of the phrase. [00:05:50] Speaker 01: in doing invalidity, he didn't apply the ordinary meaning of the phrase, intelligent agent. [00:05:55] Speaker 03: He didn't apply the district court's meeting, right? [00:05:57] Speaker 01: Yeah, didn't apply the district court meeting or any other. [00:05:59] Speaker 03: And even with- Did the district court apply its own construction in anticipation? [00:06:04] Speaker 01: No, the district court granted summary judgment based on appellee's expert's recitation of his understanding of how we were interpreting the claims for infringement. [00:06:19] Speaker 01: And there's no evidence supporting that. [00:06:21] Speaker 01: And if you look at page 25 footnote 4 of the district court's opinion, the district court recognized itself that what Appellee's expert called AIT's interpretation is more properly characterized as an anticipation interpretation. [00:06:43] Speaker 01: Because they put forth no evidence that we ever took that interpretation or were taking that interpretation. [00:06:51] Speaker 01: Appellee's own expert for anticipation said, well, I'm going to interpret it as any program that detects any changes. [00:07:01] Speaker 01: We never interpreted it that way. [00:07:03] Speaker 01: We're pointing to changes in metadata that admittedly, and not disputed for purposes of this appeal, those are changes that affect an application. [00:07:13] Speaker 01: Now, for anticipation, they say, well, and what I'm talking about, I know I'm jumping back and forth, and I apologize. [00:07:21] Speaker 01: But for POP, what they say is POP has an input field, and someone typed something in, and it's stored. [00:07:29] Speaker 01: So their expert didn't come and say, I think that's a change that affects an application. [00:07:33] Speaker 01: He said, under an anticipation interpretation, I would consider that a change that affects an application. [00:07:41] Speaker 01: And I asked him at deposition, well, how would you apply the claim language? [00:07:46] Speaker 01: Do you think it discloses automatically detecting changes that affect an application? [00:07:52] Speaker 01: And he said, I didn't consider it. [00:07:54] Speaker 01: Instead, I combined it for purposes of obviousness. [00:08:00] Speaker 03: Setting aside that question, [00:08:03] Speaker 03: of intelligent agents for a minute. [00:08:08] Speaker 03: Do you think the words changes that affect and first layer, I think it's first layer, have to be interpreted? [00:08:16] Speaker 03: Because there seems to be maybe [00:08:18] Speaker 03: two different interpretations that are being advocated by the experts. [00:08:23] Speaker 03: And I wonder whether that's something the distributor is supposed to resolve. [00:08:26] Speaker 01: Well, on the changes that affect, the experts actually agreed on the ordinary meaning. [00:08:36] Speaker 01: Salesforce's own expert testified the ordinary meaning of changes that affect indicates that the source of the change is one step removed from the claim system. [00:08:47] Speaker 01: not incorporated within it. [00:08:48] Speaker 01: So you're not just typing a change into this. [00:08:50] Speaker 01: You're not typing a change into an input field and storing it. [00:08:55] Speaker 01: That's not automatically detecting a change that affects an application. [00:08:59] Speaker 01: That's typing in someone's name. [00:09:02] Speaker 01: So I don't think there's a dispute on plane construction. [00:09:05] Speaker 03: Are there changes that affect an application interpretation or just interpretation of changes that affect? [00:09:15] Speaker 01: The district court didn't use claim construction at the basis of its decision. [00:09:20] Speaker 01: All the district court said. [00:09:22] Speaker 03: So it's just ordinary meaning? [00:09:23] Speaker 01: Yeah. [00:09:24] Speaker 01: Ordinary meaning is a question of fact and dispute. [00:09:27] Speaker 01: And they never asked for it to be construed below. [00:09:29] Speaker 01: And with respect to the third layer limitation, or the first layer, which is information about unique aspects of an application. [00:09:43] Speaker 01: Foreign invalidity, the district court granted summary judgment as a matter of law. [00:09:48] Speaker 01: The district court relied on a single paragraph of Appellee's expert report. [00:09:54] Speaker 01: That paragraph never articulates how [00:09:59] Speaker 01: The model of a car, the word Chevy, that's data stored, it says Chevy, how that would be considered or why that would be considered information that is about describing, defining an aspect of an application. [00:10:14] Speaker 01: On its face, it's not. [00:10:16] Speaker 01: And the district court adopted that testimony as true, said on summary judgment of invalidity, a reasonable juror couldn't find otherwise despite reasoned [00:10:28] Speaker 01: expert opinion from Mr. Zakovich, appellant's expert. [00:10:33] Speaker 01: Appellant's expert said that's not true. [00:10:36] Speaker 01: If I take the word a number in an Excel spreadsheet, does that number, is that unique information that is about an aspect of an application? [00:10:50] Speaker 01: Could a reasonable juror find it's not? [00:10:53] Speaker 01: And that was the question. [00:10:54] Speaker 01: And the district court simply adopted the opinions [00:10:58] Speaker 01: of a Pelley's expert and reach a conclusion. [00:11:04] Speaker 01: And that's very clear on unique aspects. [00:11:07] Speaker 01: And it's also very clear on changes that effect. [00:11:10] Speaker 01: The district court simply adopted the opinions of that Pelley's expert. [00:11:16] Speaker 01: And on changes that effect, it was particularly inappropriate because the opinions he adopted weren't only [00:11:24] Speaker 01: weren't offered how Appellee's own expert would interpret the claim language. [00:11:30] Speaker 01: He said, I wouldn't consider this to be a change that affects the application. [00:11:34] Speaker 01: But under an anticipation construction, which he made up, there's no evidence that Appellant has ever taken the position that the claim should be that broad. [00:11:46] Speaker 01: And he said, well, if I were to interpret it that way, then there would be anticipation. [00:11:51] Speaker 03: Is this the part of the report where he [00:11:55] Speaker 03: hadn't adopted or copied the report of another expert. [00:12:00] Speaker 01: Yes. [00:12:01] Speaker 01: The entire section of POP was copied from another expert. [00:12:07] Speaker 01: which they submitted during the IPR. [00:12:09] Speaker 01: They never even talked to each other before. [00:12:11] Speaker 01: He couldn't remember when it was prepared, how it was prepared. [00:12:15] Speaker 01: So his opinions weren't only conclusory. [00:12:18] Speaker 01: They weren't his opinion that he believes it should be interpreted that way, because he said, I'm implying AIT's interpretation. [00:12:27] Speaker 03: So this is a credibility issue. [00:12:28] Speaker 01: And there's also a credibility issue. [00:12:31] Speaker 01: And then most importantly, I asked this court on anticipation, or the most egregiously, I hate using that word, but claim 25 requires a metadata database. [00:12:43] Speaker 01: The district court granted summary judgment neither of Polly nor the district court even identified what the alleged metadata database is in POP. [00:12:58] Speaker 01: There is no metadata database in POP. [00:13:01] Speaker 01: They didn't identify one, and their expert didn't even address that claim. [00:13:08] Speaker 02: Counsel, you're into your rebuttal time. [00:13:10] Speaker 02: You can use it or save it. [00:13:12] Speaker 01: I'll save it. [00:13:13] Speaker 01: Thank you very much. [00:13:14] Speaker 02: All right. [00:13:26] Speaker 00: May it please the Court, Your Honors? [00:13:28] Speaker 00: The District Court was correct in construing the asserted claims to require the use of intelligent agents, and also was correct in granting summary judgment of invalidity and non-infringement. [00:13:39] Speaker 00: And I want to go to the claim construction issue with respect to intelligent agents. [00:13:44] Speaker 00: And Your Honors, the Court correctly construed automatically detecting changes that affect a particular application to require intelligent agents, [00:13:55] Speaker 00: Because the use of intelligent agents is described throughout the specification and in criticizing the very prior art that existed in the specification, [00:14:08] Speaker 00: The patentees repeatedly criticized the prior art as not having intelligent agents. [00:14:14] Speaker 00: Where is that? [00:14:15] Speaker 00: Column 7, starting at line 46, 47, after seven, six columns worth of descriptions of all the different types of regulations that are out there that are changing all the time, it says, quote, various attempts have been made to manage regulatory compliance, but no solution has been developed before [00:14:35] Speaker 00: that provides a comprehensive integrated framework for one, absorbing business changes into the application and database without affecting the integrity of the system. [00:14:45] Speaker 00: Two, automatically making application and database changes using intelligent agent routines. [00:14:53] Speaker 00: And then it goes on and down again at line 15. [00:14:56] Speaker 03: It has three also, which is a pretty long thing that it says isn't in the priority. [00:15:03] Speaker 03: When a patent specification says there's three objectives of the invention or distinguishes prior on three bases, aren't patent owners entitled to choose which of those they want to recite in their claims? [00:15:20] Speaker 03: They don't have to choose all of them, right? [00:15:22] Speaker 00: No, but this, I think, is conjunctive, the three that are listed there. [00:15:26] Speaker 00: And I think even more, when you read the very next sentence in line 61, [00:15:31] Speaker 00: These partial solutions also do not provide a, quote, closed-loop approach to identifying changes using intelligent network agents, recommending no modifications to the business content, and automatically affecting modifications in the system without the use of programmers and our programming. [00:15:53] Speaker 00: And that's only in describing the priority. [00:15:55] Speaker 00: When we look further into specification, and Your Honor talked about it in column 10, [00:16:01] Speaker 00: The specification specifically defines the invention to include intelligent agents. [00:16:08] Speaker 03: That's the fact that it says example. [00:16:11] Speaker 03: It says the following example illustrates, then it has example A, and then it says that the language about the invention begins tracking using intelligent agents is under the heading example. [00:16:24] Speaker 00: That's the only example. [00:16:25] Speaker 00: There is no other example described in the specification that [00:16:29] Speaker 00: describes using something other than intelligent agents. [00:16:31] Speaker 03: Does our case law say that when only a single embodiment is disclosed, we're supposed to read it into the claims? [00:16:37] Speaker 00: Well, I think the case law dictates that when there is criticism of the prior art for not having intelligent agents, when there is lots of references to the invention, the system, and I'm prepared to go through other examples, where [00:16:56] Speaker 00: The only thing that's described in connection with the invention is intelligent agents. [00:17:01] Speaker 00: And then on top of that, when you have, again, the only embodiment being the description of intelligent agents, I think, yes, that that's exactly when intelligent agents then should be read into the claim language. [00:17:17] Speaker 03: And if we look at them. [00:17:18] Speaker 03: Can you tell me the word intelligent agent is defined in column 20 at the top? [00:17:25] Speaker 03: very broadly. [00:17:26] Speaker 03: I just think it could be probably any software. [00:17:29] Speaker 03: Am I misunderstanding that? [00:17:31] Speaker 00: Well, I think it's defined in a couple of different places at column 20 and also column 10, and it's pretty similar. [00:17:37] Speaker 00: And it is broad, but it's the part about affecting change. [00:17:43] Speaker 00: And if we look at column 10, an intelligent agent is a specialized program that resides on a network or at a server as an applet and can make decisions. [00:17:53] Speaker 00: can make decisions and perform tasks based on predefined rules. [00:17:59] Speaker 02: Isn't there a genuine issue as to whether the deploy function is an intelligent agent? [00:18:06] Speaker 00: No, Your Honor, there is not. [00:18:08] Speaker 00: There is a complete and utter failure of description on AIT's part [00:18:14] Speaker 00: the deploy function being an intelligent agent. [00:18:16] Speaker 00: There was one paragraph in the expert report that described the deploy function. [00:18:22] Speaker 00: Not once in that expert report does the expert describe it as an intelligent agent. [00:18:27] Speaker 00: There is no analysis of the source code pointing to where in the source code decisions are being made and tasks are being performed based on predetermined or predefined rules. [00:18:40] Speaker 00: that the court evaluated the evidence. [00:18:43] Speaker 00: And again, when we look at the deposition testimony of the expert, the expert's deposition testimony is largely its conclusory. [00:18:51] Speaker 00: I mean, he never says in his report or his deposition, here are the specific reasons why it is an intelligent agent. [00:19:01] Speaker 00: And I can go. [00:19:04] Speaker 00: into more detail about that. [00:19:06] Speaker 00: But at the end of the day, what we have here is a specification that tells a person of ordinary skill over and over again that what's important about this invention is intelligent agents. [00:19:18] Speaker 00: I heard a reference to claim differentiation by counsel for AIT. [00:19:24] Speaker 00: We all know that's a presumption. [00:19:26] Speaker 00: And in this case, the presumption is overcome. [00:19:28] Speaker 00: There are plenty of this court's case law in the Tippman case and retractable. [00:19:34] Speaker 03: Do you think the prosecution history deserves any weight? [00:19:36] Speaker 03: You know, there's prosecution history that shows that [00:19:41] Speaker 03: the applicant was not interpreting their own claims as requiring intelligent agents. [00:19:46] Speaker 03: They only made that distinctive point with respect to the dependent claims. [00:19:50] Speaker 03: Does that add on to the claim differentiation argument? [00:19:55] Speaker 00: No, Your Honor, because under the Biogen case from this court, the prosecution history cannot be used to enlarge the scope of the claims, which we know from that case. [00:20:04] Speaker 00: And on top of that, [00:20:06] Speaker 00: Because they were making one argument with respect to dependent claims, I mean, it almost confirms to a person of ordinary skill in the art that that same argument would apply to the independent claims. [00:20:17] Speaker 03: Not necessarily. [00:20:18] Speaker 03: Dependent claims usually add things to independent claims. [00:20:21] Speaker 00: Well, but not in this case. [00:20:23] Speaker 00: And again, it's a presumption here. [00:20:25] Speaker 00: And I think this case, this court's case law specifically with respect to the Tippman cases retractable, we cited those [00:20:34] Speaker 00: at length in our briefs, again, I think the presumption is overcome. [00:20:38] Speaker 00: I submit that the presumption is overcome here. [00:20:41] Speaker 00: And again, with respect to the deploy function, what was apparent from the deposition testimony was, and the best quotes that they have in their briefs, again, are conclusory quotes. [00:20:56] Speaker 00: There's never any claim limitation by claim limitation [00:21:01] Speaker 00: which is what I would expect to see, describing exactly where the functionality is with quotes to the specific source code on what the deploy function is actually doing and why it's an intelligent agent. [00:21:16] Speaker 00: With respect to validity, this is one of those cases where, again, they took a very broad view on the claim construction, and they walked right into the pop reference. [00:21:29] Speaker 00: And the pop reference anticipates each and every asserted claim. [00:21:35] Speaker 03: What do we do about the fact that it looks like the district court applied a different construction for anticipation than it did for infringement? [00:21:45] Speaker 00: I don't think it applied a different construction for infringement versus anticipation. [00:21:50] Speaker 00: And as we described, what I think we have here is a situation where AIT [00:21:59] Speaker 00: It was interesting to me, but AIT proposed a broad construction of intelligent agents that under their broad construction, I think, renders POP as anticipatory. [00:22:12] Speaker 00: And again, as we submitted in connection with our briefing, if intelligent agents is required and the specific hour definition of intelligent agents, which is the narrow construction of intelligent agents is required, [00:22:27] Speaker 00: then we believe the claims are invalid as obvious between POP and Amati. [00:22:33] Speaker 00: And there's really no argument on the other side about why Amati doesn't disclose an intelligent agent. [00:22:42] Speaker 00: Their expert never contended that Amati didn't disclose an intelligent agent. [00:22:46] Speaker 00: In fact, their expert never even contended that POP didn't disclose an intelligent agent. [00:22:51] Speaker 00: So we have here, and this is exactly what the PTAB reached in terms of its decision. [00:22:56] Speaker 00: This court reversed on procedural grounds, but the merits still remain. [00:23:01] Speaker 03: The question is whether it's a genuine issue of material fact, right? [00:23:05] Speaker 03: I mean, it's different. [00:23:06] Speaker 03: The PTAB is entitled to make fact findings, right? [00:23:11] Speaker 00: You're right that the question is whether it's a genuine issue of material fact. [00:23:16] Speaker 00: And what we have here is not one or two or three different bases in order to [00:23:22] Speaker 00: to sustain and affirm the court's ruling in the lower court. [00:23:27] Speaker 00: But what we have here is a finding of non-infringement and several basis in which to invalidate the patent, not just because of POP as being anticipatory, but also that it's obvious. [00:23:42] Speaker 00: And unless your honors have any other questions with respect to any of the issues I've addressed, I'll see the rest of my time. [00:23:51] Speaker 02: Thank you, counsel. [00:23:54] Speaker 02: Mr. Defencenzo has a couple of minutes for a bottle, two and a half minutes. [00:24:01] Speaker 01: Thank you. [00:24:03] Speaker 01: We heard counsel say that the patent repeatedly distinguishes the prior art based on the lack of intelligent agents. [00:24:09] Speaker 01: There's one paragraph in the specification. [00:24:12] Speaker 01: It doesn't use the word invention. [00:24:14] Speaker 01: It doesn't say all embodiments solve these problems. [00:24:18] Speaker 01: But more importantly, as explained in the reply brief on five through six, each of the problems identified [00:24:30] Speaker 01: This refers to metadata tables, making application and database changes, and intelligent agents. [00:24:37] Speaker 01: Those are the three of the issues. [00:24:40] Speaker 01: They're all addressed in dependent claims. [00:24:42] Speaker 01: So that background should be read consistently and can be read consistently with the claims. [00:24:48] Speaker 01: Similarly, the prosecution is read consistently with the claims and consistently with the problems identified in the background. [00:24:58] Speaker 01: At the end of the day, they point to the use of the invention once in an illustrated example. [00:25:03] Speaker 01: When you consider the evidence as a whole, we submit that's not enough to overcome the ordinary meaning of the claim and read in a limitation. [00:25:14] Speaker 01: Now, with respect to the evidence on the deploy function, [00:25:17] Speaker 01: Our expert testified explicitly that it meets the descriptions in the specification, all of them. [00:25:25] Speaker 01: He provided a five-page source code description of the deploy function. [00:25:31] Speaker 01: At the end of that description, he has boxes. [00:25:35] Speaker 01: They're in our brief. [00:25:36] Speaker 01: And the appellees had never addressed them or disputed that they disclosed tasks and objectives, and more importantly, [00:25:45] Speaker 01: A reasonable juror could find his five-page description of the tasks, rules, and objectives for the deploy function supports his conclusion that the deploy function is an intelligent agent given the broad descriptions in the specification. [00:26:04] Speaker 01: And then lastly, with respect to obviousness, obviousness is only implied if [00:26:12] Speaker 01: The claim construction with respect to intelligent agents is affirmed. [00:26:17] Speaker 01: And neither the district court nor appellees have ever explained why Mr. Zakovich's reasoned opinions concerning why these two references wouldn't be combined couldn't be accepted by a juror. [00:26:33] Speaker 01: He said the intelligent agent of Amani would have no useful purpose in POP. [00:26:40] Speaker 01: Because one's going to the internet to detect documents, and the other's an input field. [00:26:45] Speaker 01: So why would you take an intelligent agent that does something completely different and just put it here? [00:26:51] Speaker 01: And he explained it rationally and reasonably. [00:26:53] Speaker 01: And more important, a reasonable juror could accept that, especially in the face of the evidence offered by a teller. [00:27:00] Speaker 01: That is all yours. [00:27:02] Speaker 02: Thank you. [00:27:02] Speaker 02: Thank you, counsel. [00:27:03] Speaker 02: We will try to be intelligent agents in deciding this case. [00:27:08] Speaker 02: The case is submitted.