From Siloes to Synapses: Reflections on EMRs, AI, and the Future of Care

From Siloes to Synapses: Why I’m Finally Hopeful About EMRs, AI, and the Future of Care

By Dr. Carlos Yu

After 35 years in family practice, you’d think I’d have seen it all. But I’ll be honest—this moment we’re living through? It feels like the beginning of something radically new. Especially when it comes to digital health.

A colleague recently asked me a few pointed questions about EMRs and how I use them. The questions were practical—but they unlocked something much bigger. So here’s my take, in blog form, on where we are, how we got here, and what’s quietly unfolding before our eyes.

1. What manual work outside of your EMR do you do that you wish it did?

Let me count the spreadsheets.

Despite decades of development, most EMRs still don’t do the things that actually make clinical life easier—like sending group invitations, automating intake, or tracking attendance for wellness programs.

I still juggle a patchwork of Google Forms, emails, PDFs, and sign-up links to do things that should be simple.

The irony? The data is already in the EMR—it just can’t do anything useful with it.

2. Do you interlink your other workflow functions with your EMR? If not, why?

Only partially—and with a lot of duct tape and goodwill.

I’ve recently started using JaneApp to manage bookings, intake, and group attendance. It’s early, but promising.

That said, most EMR integrations still feel like awkward bolt-ons.

Why?

Because most EMRs weren’t built to empower.

They were built to bill.

Limited APIs, vendor silos, and architecture that assumes the physician is the problem, not the user.

We’re doctors, not IT departments.

We need tools that just work—not ones that need a master’s in systems design.

3. How much time do interoperability issues take away from the clinic’s staff?

Let’s just say…

If I had a dollar for every piece of information manually retyped, refaxed, rescanned, or hunted down because System A couldn’t talk to System B—I’d have enough to fund a startup to fix it.

These inefficiencies chew up hours every single week for each staff member.

Worse? They burn out good people.

And for what?

4. Will the real solution arrive before I finish building my workaround?

Probably.

Honestly, by the time I get my current system stitched together, I suspect a beautifully integrated, off-the-shelf solution will already exist.

The pace of innovation is astonishing.

So I’m watching with a strange mix of awe, curiosity, and hope.

My hunch? By the end of 2025, we’ll see something elegant, intuitive, and actually designed for clinicians.

Fingers crossed.

5. Why were EMRs so siloed in the first place?

Here’s the part no one likes to talk about.

The technology to integrate healthcare data has been around.

The issue wasn’t technical.

It was political.

Siloed data served certain interest groups.

Whole layers of bureaucracy, business models—even job security—were built around fragmented systems.

Meanwhile, the public—understandably concerned about privacy—was hesitant to support full integration.

That’s changed.

The interest groups are still there. But public opinion is shifting.

People want their data to move with them now.

That shift in awareness is the real pivot.

It’s what opens the door to interoperability—not just policy or tech, but permission.

6. The AI paradigm shift is real.

Here’s what’s different now:

With AI, we don’t need perfectly structured, relational databases with neat little columns and rows.

AI can work with messy, unstructured, multi-format data—and pull out clinically relevant insights in real time.

All it needs is access.

Give it access, and AI can do the rest.

I don’t know exactly how it works (somewhere in a digital cortex, I assume), but I know this:

The old rules no longer apply.

The database now lives… somewhere inside the neural net.

(I know—sci-fi vibes. But it works.)

7. The key now? Lean in.

Some physicians are still wary of using AI.

Totally fair—it’s new, it’s fast, and it can feel like a black box.

But I’ve started using it in my practice—and honestly, it’s already changed how I work.

From scribing, to session planning, to helping patients make sense of their own patterns—AI is saving time and amplifying clarity.

I’m not trying to replace myself with a robot.

I’m just trying to work with the machine to make space for the things only humans can do: listen, connect, understand, and care.

Final Thought

We are on the edge of something transformative.

Not just in technology—but in how we think about care, connection, and collaboration.

As physicians, we don’t need to wait for perfection to begin.

We just need to start experimenting.

Because the future of healthcare isn’t just arriving…

✨ It’s learning. ✨