From Your Hosted Legacy DMS to Gallery: Seven Questions with an Expert
Davey has spent 21 years working inside legacy document management and Epic environments and leads FastLane delivery at Quoris. We sat down with him to talk about what is happening inside hosted environments right now, and why it matters for health systems trying to get to Gallery.
Why are so many health systems stuck right now?
Within the last ten years or so, third parties have started creating hosted solutions. Everyone was talking about the cloud and wanted to move to the cloud. The benefit sold to customers was that by moving to the cloud, servers would be managed so they wouldn’t have to deal with it. When the DMS keeps everything in their data center, it frees up data center costs. And while it did save customers money at first, that did not last.
Over time those costs started increasing, and the quality and support started decreasing. At this point, you’re paying more and getting less, which is why people are making the move to Epic Gallery.
There was also a point where hosted became the only option being offered for new customers. So, if your organization came onto one of these platforms after a certain year, you were going to be hosted whether you asked about it or not. A lot of organizations don’t fully realize what that means until they try to leave.
Why is data migration more complex for hosted health systems?
If your document management system is on-premise, your data lives on your network. You have the database, the servers, the connections to Epic – all of it. Getting the data moving is just a question of working within an environment you already control.
Hosted, however, is different in ways that are not always obvious until a project is underway.
Your data is not on your network. It’s at a separate facility (possibly across multiple states) inside your vendor’s infrastructure. If everything is on your network as it is with on-prem, you’re limited by your traffic. But with hosted, you’re at a completely separate location dealing with network traffic between facilities, requiring SFTP server configuration and all the coordination that comes with it. This is where timeline delays start to creep in.
There is also no simple exit path. In the past, the legacy DMS would say, “hey, we’re just going to take this drive out of our data center, we’re going to give it to you, and now you have all your data.” They’re not offering that anymore. They’re keeping everything there, and you have to say, “okay, on this date we’re done,” and then they send that drive.
That’s not a feasible option for health systems because they need to be able to migrate that data over, while still maintaining 24/7 live access to it.
Extraction needs to happen live, in parallel, without gaps in availability, while your clinical staff are actively using the same infrastructure you are trying to extract from. Managing that without disrupting end users requires a different level of operational discipline than a standard on-premise project.
Does the vendor relationship make this harder?
It can. When a vendor is managing an end-of-life platform, their resources and priorities are naturally oriented elsewhere. Their contractual obligations don’t disappear, but migration projects aren’t at the top of their queue, and that’s just the environment you’re operating in.
Support teams are often stretched thin, and the institutional knowledge that existed earlier in the platform’s lifecycle isn’t always there anymore. That’s not a criticism of the people involved. It’s a structural reality of working within a system that is winding down, and it’s worth factoring into your timeline and expectations before the project begins.
The result is that hosted customers are often managing a stalled extraction on top of an already difficult support relationship, while the clock on their legacy contract keeps running.
What goes wrong when an extraction vendor is not built for hosted environments?
The most common approach I see from other extraction vendors when they encounter a hosted environment is to ask for the data outright: request a copy of the database or the files, pull them into their own environment, transform them there, and send the output back.
The problem with that is your PHI is leaving your environment. It moves to a third-party network you do not control, processed by people whose security posture you cannot directly verify. That exposure almost always triggers a full vendor security review before the project can begin, and those reviews add time. In a project where timeline is the whole point, that time is expensive.
When my team performs extractions, we enter directly into the customer’s already-secure environment. We’re not bringing any third-party applications in, nor are we asking them to install our own custom software. No data leaves, no off-shore resources are introduced, and the client’s security perimeter does not change because we are operating inside it.
What does a hosted extraction look like once it’s underway?
Once access is in place, extraction runs continuously. For hosted environments, that means around the clock, seven days a week, because compressing the timeline is how you minimize the overlap period on a legacy contract.
At Quoris, we have an on-call schedule across our team with 24/7 support for this process. If a client reports an issue at 2:00 AM, they can call us. We’ll answer and work with them to identify the issue and determine the next steps, no matter what its origin was.
Beyond the on-call coverage, we’re checking in three times every day: first thing in the morning, midday, and before whoever’s on call goes to bed. We take that burden off the customer. They may only have one or two admins who are occupied dealing with day-to-day operations, and they’re working on transitioning to Gallery at the same time. We’re going to manage that whole extraction and conversion process for them. We’ll pull them in as we need to, but we’re going to know this project inside and out.
Throughout the project, the customer’s internal team has real-time access to our reporting dashboards showing where the extraction stands: total documents expected versus extracted, broken down by document type, file type, etc. We’re also surfacing stuff like, “hey, you’ve got XML documents in here that need an overlay or a stylesheet.” Catching that kind of thing early helps prevent reconciliation problems at go-live.
What makes the difference between a team that can do this and one that cannot?
For Quoris, it’s our breadth of knowledge that makes us unique. I’ve spent 21 years working in legacy document management and Epic environments, so I have a lot of intimate knowledge about the integrations and pain points. We also have people who’ve worked in the healthcare space with Epic and DMS consulting in multiple areas. One of our team members comes from a banking background with these DMS platforms. That range of experience is what allows us to move through complex hosted extractions without losing momentum.
Hosted extraction projects surface surprises: document types nobody inventoried, indexing structures that don’t match the project scope, integrations with ancillary systems that complicate the extraction logic. When those things surface, and they will, the question is whether the team running the project has seen enough variation to navigate without stopping the clock.
Get in the FastLane to Gallery.
Quoris’ FastLane program is built to handle hosted and on-premise environments with equal speed and expertise. We accelerate time-to-start by expediting contract turnaround, then offer kick-off within two weeks of execution. Then, we work round-the-clock to extract and convert your data for Gallery.
Our US-based team pulls from decades of DMS and Epic experience to expertly navigate both hosted and on-premise legacy environments, across any DMS
What we support: Our FastLane team handles migration from OnBase, ChartMaxx, Quanum, OpenText, M-Files, Box, and other legacy document management platforms.
If your legacy DMS is hosted and Gallery is on your roadmap, we have done this before.
Share this