Your Agreements Are a Database You Can't Query — Hiral Shah, Docusign & Sean Sodha, NVIDIA

0:01 [music] Uh, welcome everyone to our session. Um, I'm Haral. I'm a senior director of products um at DocYsine and I'm joined by Sean.

0:21 >> Hello everyone. I'm a product manager at NVIDIA. So today Sean and I are going to talk about a massive problem that every enterprise faces which is agreement data largecale agreement data. Agreements are

0:35 a big part of any relationship any B-2B organization kind of goes through day in day out and a lot of that data is captured inside that agreements and it's very critical whether it's pricing

0:48 tables a lot of it and it's all in a lot of different unstructured format and that's kind of what we're going to show is how docysine partner with Nvidia are fixing that on making that data

1:01 available readable usable for a lot of our kind of organizations. So here is kind of just a quick map of our talk today. We'll start with just the stakes. Why does this matter? Why

1:13 the scale is so large? And then we'll dive deep into the technical architecture of how we are approaching it, how we have tackled this thing, especially this document processing at

1:23 at scale. And then finally, we'll cover what we have learned from our evaluation of all of the different models we've tried for different purposes. and share our learnings with you. So why like you

1:37 know just to understand why this is a big problem when you think about docysine raise your hands how many of you have used docuign anyone who's employed probably the HR docs right so

1:48 it's a massive scale everyone uses docy for us it's a massive engineering problem as well because just look at the scale we have 1.9 million customers who are paying us and a billion users what

2:03 does that imply we process us a million agreements a day that needs to now structurize make it readable make it querable usable and you know in the past we've worked with Deote on a study and

2:16 it says that there's $2 trillion captured in this agreement negotiated value that no one capitalizes no one goes back and gets that um data back right and why it's because they have to

2:29 do a lot of human reading human reviews there's disconnected systems lot of manual workflow flows that are there. So that's kind of why docysine built AM an intelligent agreement management

2:41 platform that takes the entire like applying in an AI first way the entire agreement life cycle whether you are creating agreements genai helps a lot with that whether you're negotiating to

2:52 understanding and redlinining all the way to after signing storing and making a lot of insights from this data. So when you think about the challenges involved right in an agreement there is

3:05 the unstructured data it could be a PDF it could be a PNG and what are people wanting to do is simple questions they can't get that that is the data that is trapped inside one agreement but also

3:18 the whole corpus of millions of agreement 10 years 20 years of business has kind of put into that. So for agreements aren't flat they're also hierarchical in nature they are like you

3:31 know one agreement governs the other the other kind of does something else so you always are needing lot of things to answer this question simple thing I'm sure you all are using a lot of right

3:41 claude and all especially at your company's organization simple thing what did we contract for the total tokens right with claude no one knows that's captured inside this agreement in

3:51 different forms and fashion so you need to be extracting this data to find things but also insights and push it downstream where you're tracking doing more things and when you analyze an

4:02 enterprise contract a big set of things are captured what I call vital terms like pricing tiers the skues the information SLAs's rate cards they're all in table format now traditional

4:15 document extraction tools or a generic VM BLM completely fail here like you know we've tried we've definitely done this because they're reading text line by line which breaks a lot of that

4:27 concept within the table. A merge sells some things are not boundaried. So this makes a massive operational overhead for the downstream legal teams, procurement team, sales teams to get that queries

4:38 and get that answers done and they spend hours and hours digging through this um to even just locate a basic thing. And that's kind of where we partnered with Nvidia and leveraged a purpose-built

4:50 model like you know tool which is literally like you know making our architecture for table extraction take us to make and solve these complex use cases. We're trying with this we're

5:02 making things scalable. We can understand it with the layout but also deliver really accurate results. And to share more about how we're leveraging the Neotron, I'm going to hand it to

5:13 Sean. All right. Hello everyone. Uh so real quick on uh the Neotron retriever initiative. So for those who here knows about Neotron, you raise your hand real

5:25 quick. Awesome. Uh so Neimatron is all about building worldclass open-source models and publishing the data sets, the techniques, uh the quantization approaches, distillation approaches,

5:36 pruning approaches, every technique possible, blueprints to go with that, you name it. Um throughout the Neimatron portfolio, we have specifically Neatron Retriever, which is building embedding

5:46 models, reranking models, and document extraction models. Um so real quick here we sort of our first initiative is if you're a large scale enterprise that deals with pabytes scale data our first

5:57 initiative is how do we make sure that you find the right document given a certain query your agent sends you know set of queries to the corpus afterwards. Once you find those top five quer

6:07 documents whatever it may be then we say okay you found the right document now how do you then find the right information within the document and this is where the work with the docysteine

6:15 team has gotten really great where we've uh worked with them to build the neatron parse model to focus specifically on table extraction which is a really complicated technique um if you think

6:25 about it the number of permutations of tables are quite vast when you think about nested tables merged cells merged columns merged rows whatever it may be and that can get really really complex

6:33 and really hairy of a Uh so real quick as I mentioned before right our team is responsible for uh leading a lot of the leaderboards in the retrieval space. So Vidori V1, V2, V3,

6:44 MTB, MMT. Um so our team knows how to build world-class retrieval models given a lot of leadership uh given a lot of leaderboard winnings that we've had in the last year or so and then of course

6:55 as I mentioned before we open source everything right so we share the open source model weights the techniques and then we release with those blueprints and skills that agents can use then

7:04 afterwards um so to touch a little bit on the actual model that we are working with with docuign was the neatron parse model so when you think VLM you generally think a multi-billion

7:14 parameter model it's very heavy It's high latency. Um, this is a very small tiny C radio VLM. It's about 850 900 million parameter model. Uh, designed to kind of be that all-in-one package sort

7:28 of model where you deploy it and instead of having small, let's say, YOLO X models that do table extraction or page element extraction, whatever it may be. This is a singleshot model that you can

7:38 feed a document in and out comes the semantic formatting layouts the text uh the reading order uh the the preserved structure of the table etc. Um this can be served via the NVIDIA NIM or via the

7:50 LLM as well too. Um and so it's a tiny small model that you can use. It's not a generator. It's more of an extractor at the end of the day. Uh so real quick as well too um we

8:01 always want to make sure that we're building towards benchmarks that matter most to the enterprise space. So we want to make sure that both on the paro curve of accuracy versus performance, we're

8:10 make sure that we're going to be releasing a world-class models to the ecosystem too. So you'll see here generally is just a very standard benchmark of table extraction. I believe

8:17 this one was RD table bench and we compare some popular open source models here and then we compare how our neatron parse model does compared to that industry and we continue to kind of

8:27 strive to improve this as time goes on. >> So that I think we believe we have a demo as well. Yeah, >> just press one. Yeah. Okay, there we go. >> Let me show you how easy it is to turn

8:38 any agreement into structured usable data with agreement manager, which is a central repository of every agreement an organization has ever signed. Let's look at this. So when we look at the

8:51 agreement manager view here, you know, we have an ability to see the entire list of agreements, but also go and upload a new agreement. So I'm uploading a new order form into agreement manager.

9:03 As you can see, I can select from my computer, import from other places. The moment I select the agreement, it starts uploading and starts processing with AI. And just like that, you can see that the

9:16 jobs engine has processed it. Let's take a closer look at this agreement. So when you go into the action, you can go and browse the file. Within seconds, agreement manager has extracted a rich

9:28 set of metadata. Everything from key terms to commercial details are automatically structured, highlighted, and immediately you can jump to that section where the details are found.

9:40 Builtin goes deeper. This is where the Nvidia's model comes in that it's extracted all the structured pricing data around this agreement. It goes in breaks down these complex t tables into

9:52 order details. And as you can see, we can break it down. We can download all of this data. This is powered by the advanced parsing leveraging Nvidia's Neotron model,

10:05 turning every even dense tables into something that is instantly usable. And of course, you can take this data with you. You can see when we've downloaded into CSV how we've structured all of it

10:15 for your finance team, procurement team, even further analysis. All of it is also available through API. And that's how agreement manager has transformed agreements into actionable insights in

10:26 seconds leveraging Nvidia. So I think you know what you saw there from um a demo perspective we've tried to shortened it. It's like we have a whole repository what you see a list we

10:38 get customers which has thousand agreements to all the way millions of agreements within but the big piece is how do we understand and get that data that makes it very valuable to an end

10:49 business user right a legal person a procurement person a salesperson who's doing a lot of the deals or even a leader right like a business unit the CTO goes and asks what did we do this is

10:59 how we have making each of the things a lot more structured so we have our own proprietary agreement data model which we are structurizing each agreement but also at a whole organization level and

11:10 leveraging a lot of the NVIDIA things we've been able to do a really good job especially with all of those tables like pricing SLAs's and then make that available and then we also have a like

11:20 robust kind of search that is um on top of it so when we think about what have we learned right when you think from a neotron plus docyign we one of the biggest things for us we definitely have

11:33 done lot of different models mod for different purposes. So a purpose-built model for the job you're trying to do is a big big part of how we've been thinking about and that's kind of where

11:43 we've been able to accelerate bring things to market much faster. The second big piece around like the model efficiency. So for you know as Sean was talking about the number of parameters

11:54 yes context and stuff matters in the you know in a different environment for different things. For us the lower kind of context basically also meant lower latency lower cost to deliver the scale

12:05 that we are talking about last around the faster extraction. So um we we ran this against a lot of the other open source models when we think about how many tables can it extract per second

12:19 Neotron was 20x faster which helps us when we're talking about the millions and billions of scale that we're kind of serving for all of our customers. So a lot of it is like having that smaller

12:31 purpose-built things is the way for an enterprise as an organization to go and leverage and then serve that from an enduser perspective. Um and then what's next? So I'll let

12:43 Sean talk through those. >> Yeah. So working with the docuign team uh has been awesome so far. Uh and we're going to continue to deepen that partnership as well over the next few

12:51 months. So uh with them we started with the how do I extract as much possible information from a page and now we'll scale to how do I now find that page to begin with. Um so we'll start a little

13:00 bit with the neatron retriever uh effort and then of course we'll talk a little bit about the NVIDIA agent toolkit with them over the next few months um and then actually start scaling into into

13:10 more production scale agents then. >> Perfect. I think that's what we had. We have time for a couple questions. I went in the room. Okay. Someone there.

13:44 So just to recap for everybody if you didn't hear it was the question is right like OCR is always a thorn in the whole process so are we thinking about letting the go of that and starting from agentic

13:54 from the get-go I can talk from my perspective but so I think for us right like there are different use cases at different points in time many times if you are reactive you have a question and

14:05 you're coming some of that can can work dynamically at a smaller scale. The question is the latency when I am quering at that scale of thousands I do need to have pre-processed have

14:16 identified. So that's one. I think the second big part of the use case for us a lot of times businesses want to use this data to do a lot of downstream work. So an example is a procurement team. This

14:27 is my pricing table. I want to put it into Koopa to make sure when I'm paying that works at that time there like you know the agent is kind of helping but I can't do that on a one document by

14:36 document. That said, there is ways that we are compressing. That's kind of why Neotron worked for us is like how do you do it from a layout understanding just for that purpose but I would like let

14:46 you add. >> Yeah, I think it depends on the use case a little bit. Um I think for this specific instance, right, you have pabytes of documents that you want to be

14:55 queriable at some point, right? So you are heavy on the compute at the upfront side with all the OCR so you don't have to worry about it later on, right? I think there's some instances where

15:03 people may upload a contract to begin with for Q&A and that's a very high that's a very low latency use case, right? So you have a high throughput versus low latency use case and in that

15:11 scenario your different batch sizes, your concurrencies, your different techniques on how you process the document will be different and where you spend that compute in that cycle will be

15:19 changing between the different use cases. >> Okay, one more there. We we do a lot of like more of what I call hybrid approach and a purpose-built

15:40 for like the needs and the use cases. So from a table piece it does kind of you know do the whole layout along with extracting we still do OCR from a lot of other fields and metadata and a closet

15:51 like all of the text kind of thing. So that I think we had a architecture where we have our pipeline going through two different routes for that. Um as a follow we have a blog out there how are

16:00 we really solving this at scale across and if you look at that there's a lot of different peacemail modules and stuff together. >> Yeah one moreization

16:14 >> so we this model is currently on FP16 but there are paths towards going on to FP8 and VFP4 in the next few months as well too. What about center? Are you using the 16 or

16:26 >> we do use that and then we're also kind of using some of the older ones and that's the journey as a partnership is to kind of go tweak as you get more of the customers.

16:34 >> Yeah. So for this there are many techniques on how to improve the performance side right so quantization right so we're trying to move everyone to blackwell right so that's why NVFP4

16:42 is the big thing now um as well as multi-token generation for this it's a VLM architecture right so your encoder decoder techniques can definitely be further optimized so not right now this

16:53 model just generates one token at a time you can do multi token generation of course too so there's plenty of performance things right now we're focusing on the accuracy side like are

17:00 we adding value to the system and then from there we'll then push out that paro curve on the performance Are you guys using >> I believe they just deploy via directly

17:17 recall when you do that. >> Oh, this is just an extraction. This is now a retrieval. >> Yeah.

17:33 >> Yeah. Yeah, you're asking question. >> It's coming from that agreement data that we've kind of extracted and stored. >> So yeah, maybe I can chat with you offline and how like we our architecture

17:45 kind of works fully as well. No, we're almost coming up on time there. Um, but I think that's kind of all we have. I'm happy to hang around uh in the back with more questions. Um, and

17:58 good luck with a lot of your uh challenges with AI. So, thank you. >> Thank you. >> [applause] [music]