At its core is a new information architecture designed for secure collaboration and AI, and it introduces a host of cutting edge functionality including multi-party collaboration, AI-assisted smart matching workflows, recurring invites for loan servicing, and a redesigned template library, to name just a few.
We sat down with FileInvite CEO James Sampson to talk about why the company chose to rebuild instead of retrofit, how the new architecture changes what’s possible with AI, and what it means for lending teams.
Because you can’t retrofit your way to real collaboration. The problems our customers are solving have changed, and the old foundation wasn’t built for where they were going. When FileInvite started, document collection was mostly one lender asking one borrower for a list of files. That’s not what complex lending looks like today. An SBA 504 or 7(a) loan can involve a bank, a CDC or LSP, the borrower, multiple guarantors, a third-party lender, and outside counsel, all needing overlapping documents at the same time.
We could keep adding features on top of an architecture designed for a two-party transaction, or we could rebuild the foundation for how lending actually works. Retrofitting is faster in the short term, but every feature you bolt on makes the system more fragile and more limited. We chose the harder path. Two years later, we can do things for our customers that we simply couldn’t before, and that platforms built the old way can’t do now.
Lending is getting more documentation-intensive, not less. According to Coleman Report, the median SBA 7(a) loan now takes 21 days to fund after approval, a 31% increase since fiscal year 2024, and a lot of that comes down to more documentation, more closing conditions, and more third-party work. The SBA has introduced another round of tighter underwriting requirements for fiscal year 2027. Every one of those changes lands on the same lending teams, who are now collecting, verifying, and tracking more documents across more parties on every loan.
When we looked at where time was actually being lost, it wasn’t inside one organization’s workflow. It was in the handoffs between organizations: forwarded attachments, duplicate requests, and multiple teams working from different versions of the same loan file. At the same time, AI was becoming genuinely useful for document work, and we knew our customers would need a platform that could put it to work safely. Both of those pointed to the same answer: rebuild the foundation.
The biggest change is how information is stored. Most document platforms, including the original FileInvite, are built as a hierarchy. Everything lives inside one organization’s account: the account owns the loan file, the loan file owns the requests, and the requests own the documents. That works fine when one lender is talking to one borrower. But the moment a second organization needs to work on the same loan, the data is siloed inside someone else’s account, so the only option is to copy it, forward it, or re-request it.
We rebuilt our information architecture to decouple how information is stored from which entity it belongs to. Documents, requests, contacts, and permissions are no longer locked inside a single account’s hierarchy. That means separate organizations can securely work on the same checklist, see the same documents returned, and share the same audit trail, while each one keeps control of its own data, templates, and contacts.
That’s the difference between collaboration that’s built in and collaboration that’s bolted on. Platforms built on a hierarchy can add sharing features, but underneath, the data still belongs to one account and has to be copied to reach anyone else. A retrofit can make that look seamless, but it can’t make it work like a shared file.
The goal was never new technology for its own sake. Every architectural decision came back to one question: does this make it faster and more reliable for a lending team to get a complete, compliant loan file at the volume enterprise lenders operate at?
Document collection in complex lending has always been a team sport, but the tools haven’t treated it that way. Every deal involves multiple organizations asking the borrower for overlapping documents. The cost of that shows up as delays, rework, and compliance risk, and the borrower feels it too when they get the same tax return request from three different parties.
Invite Sharing lets separate organizations work from a single, shared document request. Everyone sees the same requests, the same documents, and the same real-time status. The originating lender stays in control of the invite, and every upload, review, and approval is captured in one timestamped audit trail. It’s built for the partnerships lending depends on: banks and CDCs on 504 loan packages, 7(a) lenders and LSPs, lead lenders and participants, brokers handing off borrower files, and lenders working with outside counsel through closing.
That was a non-negotiable for us. Collaboration only works if it includes everyone on the deal, not just the parties who happen to be customers. Invite Sharing extends to third parties not yet using FileInvite, such as attorneys, LSPs, third-party lenders, and guests. They get a simple, secure login to view the checklist and access documents from one dashboard. No new software to buy, no training required.
Most companies are adding AI to platforms that were never designed for it. AI is only as good as the data it can reach, and in a hierarchical system, that data is fragmented: split across accounts, trapped in email attachments, and disconnected from the rest of the loan file. You can put an AI feature on top of that, but it’s working with one piece of the picture.
Our rebuild changed that. Because information is now structured and decoupled, AI can work across the entire loan file, across every party contributing to it, and across the full lifecycle from origination through servicing. That’s what we mean when we say the platform is fit for purpose for AI. We’re not attaching AI to an old system and hoping it helps. We built the foundation so AI can do meaningful work for our customers, safely, in a way we couldn’t before.
We’re also practical about AI. It has to remove real work from someone’s day. AI-assisted smart matching is where that starts. It uses document classification and data extraction to identify what a borrower submitted, automatically mark it as received, and route it to the right place. That means faster processing, fewer mismatches, and a cleaner loan file from submission through review, without a processor opening every file to figure out what it is.
But smart matching is just the start. The same foundation lets us move toward checking whether a loan file is complete before it reaches underwriting, validating documents like bank statements against a lender’s criteria, and pre-filling data so teams aren’t re-keying information from one system into another. Those are the jobs that eat lending teams’ days, and they’re exactly where AI earns its place.
We take that caution seriously, because our customers have to. Most of the lenders and CDCs we work with now have formal AI policies, and they should. In a regulated industry, AI has to be transparent, controllable, and accountable, or it doesn’t belong in the workflow.
So we’ve built it that way. AI in FileInvite supports people’s decisions; it doesn’t replace them. Your team still reviews and approves what goes into the loan file, and every AI action is captured in the audit trail. We don’t use customer data to train AI models, and we don’t share it with third-party AI model providers. Customers own their documents and the AI outputs generated from them. Our AI terms are public, so a lender’s compliance team can see exactly what we’ve committed to.
The goal isn’t AI for its own sake. It’s giving lending teams back hours of manual work without asking them to compromise on compliance.
Because the document lifecycle doesn’t end at closing. Servicing teams request the same documents from the same borrowers on the same schedule every year such as financial statements, tax returns, and insurance certificates. Even teams with a legacy tickler system still spend hours on follow-up calls and emails.
Recurring invites with full automation of the chasing and due date follow-up automate those requests on a calendar schedule, keep loans compliant, and create an immutable communication log. It’s part of a bigger shift for us: FileInvite is no longer just a document collection tool. It’s the operational backbone of the full lending document lifecycle, from origination and underwriting through closing and servicing, and that connected record is exactly what makes AI more useful over the life of the loan.
Every time we show the new platform, we hear the same things: it’s easier to use, it saves real time, and it finally bridges the gap with their borrowers in a way their old process never did. That tells us we’re building something that genuinely makes life easier for our customers and the people they serve. One newly migrated customer told us that the new enterprise platform is going to change the industry and that we are “batting a thousand!”
The feedback that means the most comes from teams running real loan volume. Mark Abell, SVP and SBA Division Leader at PB&T Bank, told us the ability to ask or answer a question about a specific document, with the comment embedded right alongside it, has been spectacular, and that the move to the new platform was smooth for his team and their borrowers. That’s exactly the outcome we were designing for: less friction between people working on the same loan.
Existing customers fear migration pain more than they want new features, and they’re right. Their lending programs can’t stop for a software change. That’s why we’re migrating customers with dedicated support and keeping disruption to a minimum. A typical migration is completed within 30 days and includes training and set-up.
We also rebuilt the template library from the ground up, organized by loan type with categories and subcategories. Teams can select multiple templates at once to assemble a complete loan package, with role-based access controls, so getting set up on the new platform is faster than it was on the old one.
Our standards didn’t change; the foundation got stronger. FileInvite continues to uphold the highest levels of security and compliance including SOC 2 Type II and GDPR compliance, with encryption in transit and at rest, role-based access controls, multi-factor authentication, and full timestamped audit trails for every upload, review, and approval. Lenders handle some of the most sensitive personal and financial data there is. The platform must stand up to regulatory and audit scrutiny every day.
Invite Sharing and smart matching are some of the first capabilities built on the new foundation, not the last. We’ll keep building where collaboration and AI together remove the most manual work from lending teams. The broader goal is simple: every party involved in a loan to stay in sync as the loan progresses. All parties should work from the same checklist, know what’s outstanding, and trust that the documents are secure and the right people have the right access. This launch is about more than new technology. It’s about removing the friction between lenders and the partners they close deals with.
See the new FileInvite
About FileInvite
FileInvite is the secure document collection platform purpose-built for complex, high-collaboration lending and financial services workflows. From SBA lenders and CDCs to commercial real estate teams, asset-based lenders, and legal counsel, FileInvite replaces email-based document chasing with one structured, automated, fully auditable workflow, giving lending teams the speed, visibility, compliance readiness, and borrower experience their clients and regulators demand. FileInvite enables lenders to collect documents 64% faster, process 6x more loans, and save an average of 10.5 hours per user per week.
For more information or to request a demonstration, visit fileinvite.com.
Media Contact: Lisa Thiel, Director of Marketing, l.thiel@fileinvite.com, 269-806-8489