Redox Chief Product Officer Rachel Witalec recently joined Zak Pines, VP of Partnerships at Intellistack, for an episode of Partner Pulse, a series exploring how technology partners are solving healthcare’s hardest integration challenges. Intellistack published their take here. This piece goes deeper on what Rachel covered: why the data infrastructure layer determines whether activation tools actually work in production, and what that means for the healthcare IT teams evaluating them.
Getting data out of an EHR is a solved problem. Getting it into the right hands, in the right format, at the right moment, without someone manually carrying it across the gap; that’s the part most organizations are still figuring out.
The instinct is to solve for a better activation layer: a smarter workflow tool, a no-code platform, an AI-powered interface. And that instinct is right, but only halfway. A great workflow tool built on top of inconsistent data will still break. The infrastructure underneath has to be solid first.
For healthcare IT leaders evaluating AI and workflow automation tools, that distinction matters more than most vendor conversations acknowledge.
The scale-up is where most activation projects quietly fall apart
Most data infrastructure was built to solve a movement problem: get the data, move the data, done. As Rachel has written, movement isn’t the hard part anymore. Readiness is.
Organizations prove out a workflow in a controlled setting, the scale-up begins, and then it stalls; not because the workflow tool failed, but because the data arriving at it wasn’t consistent enough to be dependable in production. Epic behaves differently from Athena. Athena behaves differently from eCW. Each has its own field quirks, HL7 extensions, and patient matching logic.
When Intellistack Streamline receives data from an EHR through Redox, none of that variability lands on Streamline’s doorstep. Handling it is Redox’s job so Streamline can focus entirely on its own.
“We’ve worn those scars on Epic and Athena and eCW,” Rachel said in the conversation. That institutional knowledge — knowing how these systems actually behave, not just how they’re documented — travels with every implementation and doesn’t have to be rebuilt from scratch each time.
How Redox and Intellistack show up together for customers
When customers work with Redox and Intellistack together, the model is straightforward: each layer does what it does best, and the teams closest to the work are the ones running the processes.
Redox shows up on implementation calls not as a behind-the-scenes vendor, but as a resource who can explain how data will move, what to expect from a given EHR environment, and where the edge cases live. The goal is for customers to feel confident before they’ve written a single line of workflow logic, not discover the hard parts six months into a deployment.
“We’re seeing a need for more Intellistacks in the market,” Rachel said. This looks like vendors who can take reliable data and build real workflows on top of it, without putting the integration burden back on clinical teams.
The question worth asking before you pick your activation layer
For CIOs and healthcare IT leaders evaluating AI and workflow automation tools, the practical question isn’t which platform does everything. It’s which infrastructure layer you can trust to handle the data problem. The end goal: your stack can perform the way it was designed to when it leaves the demo environment and hits production.
The organizations making real progress aren’t waiting for a single platform to solve everything. They’re assembling stacks where each layer does its job, and trusts that the layer underneath is doing its job too.
That’s the foundation. Everything above it gets to assume it’s working.
Watch the full Partner Pulse conversation with Rachel Witalec and Zak Pines here. Read Intellistack’s companion piece here.
Related reading: Stop blaming the models: the AI is ready. The data isn’t. and The software you use every day is becoming invisible.