Inside the M&A virtual data room: secure, smart and AI-driven

• How enterprise AI is reshaping the M&A data room.
• An Intralinks view on secure, smart, AI-driven diligence.

A vendor leader's perspective on where AI-driven data rooms are heading.

Welcome to Transform Now, the podcast brought to you by SS&C Blue Prism.

Enterprise AI isn't just a tool, it's the heartbeat of smart organizations like yours, helping to reinvent work to deliver higher growth, more customers, and better experiences.

Join us as we discover how AI-powered work orchestration and automation is making things better for everyone.

Customers, employees, shareholders, and most importantly, you.

Let's get into the show.

Hello everyone, Brad Herston here.

Welcome to Transform Now.

My guest today is Richard Anstey, head of customer experience and automation at SS&C Intralinks.

Intralinks is an intra-enterprise content management and collaboration solution that plays a major role in the global transaction markets.

Welcome Richard.

Why don't you tell everyone a little bit more about yourself?

Hi, Brad.

Yeah, so as you said, Richard Anstey, my broad title is Head of Customer Experience here at Intralinks.

Been largely in product roles before that, so I ran a global product management for a big company called Open Text.

I was also chief architect of Open Text for some time.

Moved into Intralinks really because I wanted to join the world of SaaS.

I was fascinated by the growth of a exciting fintech in the mergers and acquisition space having been involved heavily in M&A in my previous roles.

So yeah, we've been through quite an evolution at Intralinks since I joined, which was 12 years ago.

Been through several exits and finally find ourselves here as a business unit of SS&C, which is a fantastic place to be.

Well, this is the first time we've actually had IntroLinks on our podcast.

So I'm really excited to have you here.

IntroLinks being a sister company to Blue Prism under the Decency Enterprise.

So Richard IntroLinks has led the market for decades in secure document sharing and deal management.

The term virtual data room gets used quite a lot to describe what you do.

Can you give us an overview of interlinks and your value proposition to the market.

Sure.

Well, let's start with that term.

So virtual data room is the progression from what used to be called a physical data room.

So if you were running a large M&A transaction back in the early 90s, you would have booked space probably in a lawyer's office and packed it full of all the documents and credentials associated with a large asset that was being sold.

People would book time that potential buyers of that asset would book time in that physical room to go through all of that documentation.

Clearly they weren't allowed to take it away because all proprietary and confidential.

There would be a high powered secretary type role who would take careful note of what you looked at, how long you spent looking at each part of that content set.

They'd also record carefully any questions that you asked about the asset in order to judge how interested a potential buyer might be.

And Intralinks was really the pioneer in recreating that experience through a SaaS application.

So rather than needing a physical space, suddenly you could do all of that online.

But bear in mind, that was back in the late 90s, which was, if you think about it, before most people would even use a credit card online.

And Intralinks was facilitating multi-billion dollar transactions.

through that medium of a web browser and a website.

And clearly the technology has moved on significantly since then.

But we are the market leader in facilitating M&A transactions through a web application.

We're also a leader in the alternative investment space.

So we have a purpose-built platform.

that covers fund management, fund reporting, investor reporting, fundraising, and all of those things around the alternative investment space.

And our new platform really is a end-to-end set of capabilities for supporting everything that goes around an M&A transaction all the way from early marketing of a deal, creating deal teasers, using powerful AI to facilitate that set up preparation of all of that content space.

And then...

using AI to accelerate the due diligence process all the way through to closing and archiving, having online compliance archives of a transaction and even a cross-deal analytics over a set of closed transactions all the way through to post-merger integration.

So we've built really an end-to-end platform set of capabilities that can really help that community of dealmakers to get their jobs done smoothly, efficiency and with the utmost of security, which is really where the Intralling's brand plays is in this highly secure facilitation of sharing of information that needs to pass across and between organizations, but is often market moving data that cannot be seen externally.

You can often have multiple parties to a deal, must all look at the same information, but must not know of each other's existence.

And so element of wrapping high secure processes around confidential information sharing is really the sweet spot of what we do.

And we've applied that in a set of different areas across financial services.

Great.

Give us a sense for the volume of activity that you support.

How many transactions run through in trilings each year?

Yeah, so it will be somewhere in the region of 10 to 15,000 M&A transactions that will go through on, well, which will at least start in any, any year.

So, yeah, if you think of our contract volume, it's kind of in excess of that amount that totals something like, well, somewhere up to 100,000 invoices that have to be processed around those things.

So it's a high scale operation.

We've used automation to make those processes.

more efficient in the way that we go about delivering our service to customers.

Absolutely.

You start talking about incredible volumes of documents and automation usually comes right to the surface of that conversation.

Richard, you've mentioned some of the reasons why interlinks exist and some of the things that you facilitate.

Maybe you could elaborate a little bit further on some of the biggest challenges that your customers face today.

in managing sensitive financial information that is really the crux of these M&A transactions.

What are those and then how does interlinks address those?

Our customer base really covers the vast majority, I think with 99% of the 41,000 companies, all really major global banks and financial services institutions have been customers.

And they're obviously under a lot of regulation around the management of information and how to look after content that may be long to them or may be long to their clients.

And so those institutions go through quite a lot of diligence even before signing off on using a platform like ours.

Of course, they have a lot of regulation and process around how to handle information that stays within the walls of that organization.

When it comes to sharing confidential data, market moving data through the medium of the internet, things really ramp up in terms of the diligence they have to go through and the testing they have to go through to become confident of using a service like ours.

And so, you know, we are under.

regular audit by really all of the major global banks to ensure that we have the processes that you need in order to run a service that can handle that type of information and then volume and give confidence to that type of customer that they're not doing anything dangerous that we've got it under control.

Right.

And how has AI been utilized to extend the capabilities of your platform?

Could you talk about that?

Where it's being leveraged?

Sure.

So we've been using AI for a long time.

We were using LLMs incidentally before they became kind of popular with chat GBT.

So we were using LLMs in the first instance for redaction.

I talked about that process of preparing content before you will put it in front of potential buyers of an asset.

Well, clearly redaction plays an important role there because maybe the potential buyers of that asset are competitors of that asset.

You don't want to share sensitive, competitive information and you certainly don't want to share personally identifiable information until you absolutely have to.

And so, a process of redacting a huge content set prior to an M&A transaction can be, well, it can take a lot of time and it can be quite expensive if you have to employ paralegals to go through that effort by hand.

of marking up documents.

So we'll be using AI from a very early stage to identify entities as they appear inside a text, things that perhaps should be redacted, suggesting those things and then allowing a user to simply say, yep, that looks right.

And then the redaction gets applied across the content set for when it's consumed by the end user.

Now, since then, we've been using AI quite a lot in the customer experience side.

So we have really powerful chatbot.

for our support interface that happens from, well actually it happens across a set of media from our website through to our knowledge center and through to inside the product itself, where people need help or guidance on how to do something or if they get stuck.

Obviously we have 24 seven support but sometimes you can get an even faster answer from an AI that understands your context, what you're trying to do and has ingested a whole set of existing documentation and as learned from past support tickets of what people struggle with.

But really the latest innovation that we have is the thing that we call LINK, which is something that can understand the content itself as it sits within our repository.

So as a potential buyer of an asset, you have a set of permissions inside a virtual data room to see certain content.

Well, you can ask highly detailed questions to our link AI, things like, well, tell me the EBITDA from three years ago, and how has it changed over the period since then?

And the AI has that level of understanding of the document set that it can answer questions like that in full without that content ever leaving the secure repository in which it sits.

So we've constructed our own internal AI services that can deliver that level of understanding of the content set that an individual user has permission to and really get involved in very detailed financial questions or diligent style questions about the content set itself.

And really we're way ahead of the game on that, of really leveraging our own AI technology or at least the services that we build and operate entirely within our kind of walled garden.

So we're not shifting content off to open AI.

or sending it off to some other public service through the internet.

All of that knowledge is really held within the secure data room that gets audited by the major global banks, as I described previously.

Yeah.

It's fascinating to hear how you guys have used AI for so long.

And as you said, you were using large language models before they were even a thing, before Chad GPT made them common knowledge to everyone around the world.

When you start using LLMs to understand the content that's within the virtual data rooms, but also to do things like redaction, data security and privacy really is front and center.

It becomes one of the key things.

How has your team managed that carefully and made sure that your customers are comfortable with that?

And they have complete confidence.

and what interlinks is providing in the midst of this M&A transaction.

Yeah, so we've spent a huge amount of effort, you can imagine the complexity of being able to narrow down a knowledge set to just the things that a particular end user has permission to, because that's not normally the way that AI works.

It normally tries to ingest everything it can possibly find and then construct a model around all of that content, but we need to be high irrestrictive.

based on the permissions of the individual user that's engaging at that point in time.

So that's a significant part of the development effort that's gone into the evolution of the platform.

I should say as well that one of the requirements that we have is also kind of almost the opposite, which is, well, we have some organizations who say, well, we don't believe in this stuff just yet.

We think it's dangerous.

We'd like you to switch it off.

And so having the ability also in a fine grained way to be able to say, no, well, this organization doesn't want this to happen.

So even though there's kind of a large repository, maybe some participants to the deal do want to use LLMs as part of their process to run diligence.

Some organizations don't and you need to have that flexibility as well to be able to say to kind of wall off sections or parts of the capability so that an organization can take that choice based on their their level of comfort with the technology.

Let's turn our attention to the internal automation effort that has gone on across SS&C.

We mentioned your title at the beginning of this podcast that you were the head of customer experience and automation.

This is probably the first time I've heard those two mentioned in the same title.

But that's awesome that this has become a role for you as well within interlinks.

Talk about what this effort has looked like for interlinks and how it's helped you take things to the next level and really transform the way that you operate.

Love to hear just your perspective on that.

Okay.

So firstly, I understand that those two words don't naturally fit together, but there is logic behind it in the sense that we define customer experience to be really very broad indeed.

So the customer experience is really everything that the customer sees of the company.

So they see all of our external processes, whether that's going through our marketing, through our website, trying to discover what our products are all about, whether it's communicating, negotiating a quote, getting a contract, every lease of paper.

or email that we put in front of the customer is part of the customer experience all the way through to invoicing in our collections.

And so customer experience isn't the user experience that somebody has when they're using the product.

It's what it's like to be a customer going all the way through the journey.

And some of those people on the customer side don't ever even touch the product.

So the people who sign the contract or the people who receive the invoice, they probably not.

users of the product itself, but nevertheless the way that we run our internal processes can make a big difference to whether they leave with a positive experience of having worked with us or a more negative one.

So investor experiences define very broadly really we all trust cross-functional change will affect what it's like to be a customer and so program management office for example reports into me because we want to bring the customer perspective into the way that we evolve the company over time.

So maybe clarify why customer experience and automation are actually quite natural bet follows to be under the same under the same.

When you put it like that, it makes total sense.

I got to admit, so I stand corrected.

So let me give you a more real example of that, though.

I mentioned the huge volume of contracts that we do.

We have sometimes a very short sales cycle.

We can hear about a deal and deploy software or deploy a space within the same hour.

So the sales cycle can be incredibly short and every step in the way, in the process of contracting with us can be a frustration inside these investment banks when the VP says, yeah, we've got the mandate, we're going to go and operate this deal.

I need a data room in the next hour.

There's a team of people who are running around trying to figure out how to make that happen.

They have to go through all of their discovery, which data room provider shall we use and how will we get this thing up and running because the VP says go.

So everything that we do in that process contributes to whether the the end customer is kind of happy or frustrated and whether we can get this deal done.

So just to get to this hard example, we typically send out all of our quotes and contracts that they would go out through a doffcue sign process, but still, unfortunately, very few customers are ready to actually complete and sign those contracts within an electronic environment.

So many of them go through paper.

And often they're uploaded back into the docusyne-tucked environment.

But that means that we still need to review them because what if somebody's crossed something out?

What if somebody's kind of red-lined?

We can't just automate that blindly.

So what happens with every single contract that comes back into the system is that we actually load it into an intralink 6j.

And the reason that we do that, with a blue prism process, by the way, so blue prism will take the inbound contract.

put it into an interlinks exchange and then our own AI will read that contract and say, well, what's on it?

What are the terms and conditions?

Who are the parties?

It'll write that back into a JSON format into the exchange itself because it's done its extraction of the detail of the contract.

The blue prison digital worker will then pick that up and go and look back into Salesforce and see what was on the quote that we sent out.

and do a cross match.

And he's the same.

Did anybody change anything?

Has anybody redlined or kind of taken a zero off the price or added one on that would be nice.

Unlikely.

But we do that checking in an automated way so that the human element of contract review doesn't have to be read the whole thing from start to finish.

If the combination of the work that the Blue Prism Digital Worker has done and the AI response from intro links, if that matches, that's a straight through approval.

There's no need to go read that contract.

If the AI says, yeah, something a little bit different here or the blooper is an agent looks and the match doesn't quite work out, then a human will do it.

But obviously, that has time zones involved.

We don't really want to wait another day for that contract to get counter signed before we can start the project.

So there's just an example of taking a process and really enhancing the productivity of our human workers in these processes.

We're not using automation in order to reduce the workforce.

We're making them more productive by adding this combination of digital worker, our own AI capabilities, and really streamlining the process for the benefit of the customer as well as ourselves.

So the customer gets a kind of straight through, yeah, I've submitted the contract.

Of course, you can deliver the software to me so I can get started with my deal.

In addition to what you've already done in this effort, is this now kind of an ongoing thing for interlinked?

You have a lot more automations internally within your operations that you plan to execute going forward?

Yeah, so we've built a small team.

We've got a team in India who are very proficient with BluPrism itself and running automations.

We don't, by the way, define automation as only being RPA.

We have other types of automation that happen through integration of different systems or through process improvements.

And so we look at automation more holistically than just RPA or just AI or agentic processes.

But yeah, absolutely it's an ongoing process.

We have an intake form.

We have a backlog of proposed automations that have come in from all over the company, really, of suggestions where Yeah, well, this bit of work is a bit boring.

We think maybe we could automate it and we have people who will go and investigate, do some systems analysis, see whether that is feasible and then feed that into a process of creating digital workers to do that.

Now, Intralinks is not probably the classic target for a blue prism type company in the sense that we're at kind of state-of-the-art SaaS operation.

We don't have rooms full of people everywhere, anywhere in fact.

We're all doing kind of repeated tasks or you know, managing large volumes of paper, everything's electronic.

There's a lot of things that have been automated through systems already.

And so really we focus on where we can make humans more productive by adding just a little bit, maybe taking the first step or feeding information, the right information at the right time.

to the right employee in order to get a job done quicker.

I'll give you an example of that.

So we send out close to 100,000 invoices in a year.

Obviously, there's turnover of staff in our customers.

So sometimes those invoices bounce back.

Now, you send an invoice and it bounces back.

Well, you don't really want the collections team to be Calling around and saying why haven't you paid the invoice because actually we have information to say that it didn't ever get to where it was supposed to be so we have a digital worker that will monitor the bounced invoices When it when it finds a bounced invoice it will go into sales force it'll look up and see well Who's the who's the sales rep who's responsible for that account?

It'll construct an email or a task for that salesperson.

Hey, you need to contact this organization.

Clearly the billing contact has changed or left.

This is what to do when you found the new contact.

We'll notify the accounts receivable team not to go chasing and not to have the ticker running on bad debt and so on for that invoice to be paid.

So we take one small piece of information which is a balanced invoice and then the digital worker is able to go through and run all of those.

like boring things that you would otherwise have to do to undown.

Well, how do we correct this fault?

How do we make it better?

I can give you another one as well, which is that we send out a lot of customer surveys from Intralinks.

We clearly have this transactional business.

People have a data room to complete a project.

And when a project finishes, we really want to know what people thought about their experience.

Obviously with my customer experience, hat on.

That's how we make the company better is by understanding what the customer thought about this.

Now, when those surveys come back in, we have an automation that reads the survey.

If there's a textual element where somebody's filled in something that they really wanted to tell us, that is read by an AI that then decides what should we do?

How do we categorize this?

Is this a product thing?

Is it a pricing thing?

Is it a service or support thing?

Is it a kind of reliability thing?

And it categorizes the feedback and then makes sure that we let the right person know inside the company that we got this feedback.

This is the deal that it was associated with that it'll go and look up in Salesforce or in GainSight.

It'll make sure that the salesperson knows about that feedback and the CSM and the business unit leaves, by the way, our co-ceos.

They read every single one of these comments that comes back in.

So the automation will put that and wrap it up nicely into a team's message for leaders in the company and in the leadership team to see the feedback that came in.

So we kind of go from an automated process that generates surveys out to people who've completed a deal on the platform.

When we get that feedback, we make sure it goes to the right place.

so that we can learn what the customer thought about the experience that they had and as a result, improve the company to make sure that if it was negative that that doesn't happen again.

Yeah.

Those are such great examples, Richard.

Thank you for sharing those.

What I love about each of them is the value internally is so clear, but also the value to your end customer and how they benefit from the way that you've applied automation and AI.

the deal process overall, what are your thoughts and maybe predictions about how that's going to continue to change now that technology innovation is moving so quickly and AI is becoming central to almost everything we do?

Yeah, well, it's great to be at the forefront of that because we really do feel that we're revolutionizing dealmaking through the use of this technology.

If you take a process like due diligence, That's a highly information processing intensive job that often takes a lot of time of a lot of expensive people.

Interestingly, the investment bankers or the advisors who are often running these things, they're not billing that as time.

They're taking a success fee when the transaction completes.

So if we can beat up.

shorten the time that it takes to run due diligence.

There's a huge value in that to the clients who use our platform above a more traditional kind of file sharing platform that really just distributes content to the right place.

And so I think what we're going to see is that this next generation of VDR products that we're at the forefront of, they're going to transform the way that people think about, well, how do we get this job done?

How do we process the vast set of information that we've been given about this asset to figure out if we think we want it and if the price is right for it?

So being able to ask those detailed questions of a content set rather than reading 25, 90 page documents to be able to ask in a single question and rely on the answer that you get or follow up.

and a double check, cross check with the references that the AI can give you about where it found that piece of information.

That has this capacity to really speed up the time that it takes to do those things, which as I said, really has a significant value in the marketplace.

Deals get done faster, money is saved, time is saved, and we should accelerate that whole process.

Yeah.

Given what interlinks does, you have a unique...

view of the market and the state of M&A.

And you mentioned some of the volumes that you're supporting, which are really astronomical.

Maybe a final thought on just the state of the market and any trends that you guys are picking up on in the M&A world.

Yes, so M&A is interesting in that when the market is booming, there's lots of money around and people buy things.

And when the market is down, people sell things, right?

And so it has this kind of sometimes it is kind of somewhat immune to the ups and downs of the market.

The thing that often slows down M&A is the U word uncertainty.

And so I think that, you know, the situation that we're in politically in the US at the moment and that kind of lack of clarity on, well, how do I value this asset if I don't know what the tariffs are going to be, that kind of being.

can really slow down.

So we have a lot of deals.

We see a growing set of deals that are being prepared right now.

We see slightly less that are actually launching out to buyers.

So there's a little bit of, okay, we get ready, we'll get ready.

And then as we know what's going to happen, we'll go launch.

So there's some hesitation, some kind of slowing down due to the uncertainty of launch, but no slowdown in that.

activity that you go through in order to prepare and get ready for an M&A deal.

So generally, we're quite comfortable when the market is going up.

Everything's great.

People got money, they go buy things.

And then when things are going in the opposite direction, the people divest, people sell, and all of those things either way are M&A transactions.

Yeah, I can imagine.

Well, Richard, thank you so much.

for jumping on the podcast with me today.

Interlinks is a fascinating company and your value to the market is very compelling and it's just neat to hear your perspective on all things M&A but also to hear how you've embraced automation and AI internally and really make some significant improvements to the way you go to market as well.

So thank you for being here and I wish you the very best.

Well great to be here and thank you for inviting me.