There is a question that comes up in almost every orthodontic practice, usually around 4pm on a Tuesday when someone is trying to wrap up for the day: "How many new patients did we actually start this month?"
The answer exists. It is sitting right there in the database. But getting to it usually means opening a report builder, picking the right date range, choosing which filters to apply, waiting for it to load, and then realising you picked the wrong report type and starting again.
That is the problem Sparkie was built to solve.
So what is Sparkie, exactly?
Sparkie is the AI assistant built into Lightning. You type a question in plain English and it gives you the answer, pulled directly from your practice database. No menus, no report builders, no exporting to Excel.
Want to know how many patients are due for retainer checks this week? Just ask. Curious about which provider has the most availability tomorrow morning? Type it in. Need a quick count of active Invisalign cases? Takes about three seconds.
It works because Sparkie understands your database structure and orthodontic terminology. When you say "braces starts this quarter," it knows exactly what that means and where to find it.
"It is the difference between 'I will check that later' and 'let me look that up right now.'"
Why this matters for your day
Most practice management software treats reporting as a separate task. You leave what you are doing, navigate to a reports section, build the query, and come back with the number. By the time you have the answer, you have lost five minutes and forgotten why you needed it.
Sparkie sits inside Lightning and is always available. Whether you are on the dashboard, looking at a patient record, or reviewing the schedule, you can pull up Sparkie and ask your question right there. The answer comes back in seconds, formatted clearly, with the data broken down so you can actually use it.
That shift sounds small, but it changes how quickly decisions get made in a busy practice.
Built on your real data
This is not a chatbot that gives you generic advice. Sparkie connects directly to your practice database and writes queries against your actual records. When it tells you there are 18 new patients this month, that number comes from the same data your front desk sees every day.
It understands orthodontic terminology out of the box. You do not need to learn special commands or remember field names. "Patients overdue for their next appointment," "revenue from clear aligners this quarter," "which days are we busiest on" are all questions Sparkie can handle without any setup.
And because it runs against live data, the answers are always current. No stale reports, no scheduled exports, no waiting until the end of the month to see where things stand.
The kind of questions practices are asking
In the practices already using Sparkie, the most common questions tend to be practical, everyday things:
- How many new patients did we see this week?
- Which provider has openings on Thursday?
- What is our Invisalign conversion rate this quarter?
- How many patients are overdue for their next visit?
- Show me today's schedule for Dr. Williams
- What was our revenue last month compared to the month before?
These are not complex analytics questions. They are the things people ask each other across the front desk five times a day. Sparkie just makes sure the answer is accurate and instant.
Transparency built in
One thing that sets Sparkie apart from a basic chatbot is that you can see its reasoning. There is a "thinking" toggle that lets you watch how Sparkie interprets your question, what database tables it looks at, and what query it runs. If something looks off, you can see exactly where the logic went and ask the question differently.
This is not about making things more complicated. It is about trust. When your practice decisions depend on a number, you want to know where that number came from. Sparkie shows its work.
Try it yourself
Sparkie is part of every Lightning installation. If you are already on Lightning, it is waiting for you in the sidebar. If you are not, the best way to see it in action is to book a demo and ask it your own questions with your own data context in mind.
Because the best way to evaluate an AI assistant is not to read about it. It is to throw your hardest Tuesday-afternoon question at it and see what comes back.
