Why Pearl Is Taking a Different Path with AI Tutor Tools
By John Failla, CEO of Pearl Education
Over the last few months, we’ve been quietly rolling out our first AI tools at Pearl.
Not with a big launch or with flashy demos. But in working sessions with a small group of partners — states, districts, and organizations who are already deep in the work of running tutoring programs at scale.
What’s surprised some folks is that the first thing we walk partners through isn’t the product.
It’s our philosophy.
AI is a blank canvas, and without a clear philosophy of execution and impact, it’s hard to know why we are building, where to start or just as importantly, where not to go.
Before getting into that philosophy, it’s important to start with our “why” and of course, security.
We’re Uniquely Positioned to Do This — and We Don’t Take That Lightly
Pearl sits on one of the largest and most diverse datasets in all of tutoring.
That data spans:
- Student demographics
- Attendance and engagement
- Dosage and consistency
- Program design across models
- Outcomes across district-run, vendor-run, and hybrid programs
Because Pearl operates as the centralized data repository for the nation’s largest tutoring initiatives, we’re able to see patterns across time, geography, demographics and delivery models that most tools never touch. That scale and diversity matter when you’re thinking about AI.
It allows us to ask better questions, build more responsible systems, and — importantly — avoid overclaiming what AI can do. With tens of thousands of tutoring sessions happening in Pearl every week, this dataset continues to grow exponentially, which is also why security came before anything else.
Security Came First. Full Stop.
Working with states, districts, and student data means there is zero room for shortcuts.
Security, privacy, and data governance were non-negotiables for us — and they were addressed before any AI experimentation began.
AI only works in education if partners trust:
- How data is stored
- How it’s accessed
- How it’s used — and how it’s not used
Once that foundation was locked, the real question became clear: where do we start.
That’s where our philosophy comes in. It isn’t new — it’s a refinement of Pearl’s culture, one that continues to evolve while staying true to our roots: enhancing relationships, not replacing them.
Pillar #1: Enhance Relationships — Don’t Replace Them
I know this phrase gets thrown around a lot now. But Pearl didn’t start as an AI company — or even a tech-first company.
When I started Pearl almost a decade ago, it was a relationship-based tutoring company. Our focus was simple: connecting students to tutors with whom they could build deep relationships. That hasn’t changed, and it’s not going to.
I didn’t start Pearl to automate people.I started it to make it easier for everyone in a student’s academic journey — tutors, caregivers, coaches, administrators — to actually show up and play their role.
Tutoring works because of trust, consistency, and human connection.
AI’s job isn’t to replace that; it’s to remove friction around it.
If a tool weakens human connection, it doesn’t belong at Pearl — no matter how impressive the technology is. That principle acts as a hard constraint on everything we build.
Pillar #2: Ground Everything in Evidence (and Show the Work)
The biggest hurdle AI has to cross in education isn’t capability. It’s trust.
Educators are right to ask:
- Is this actually correct?
- Where did this come from?
- Would I stand behind this with a student?
We believe AI should help people access the incredible amount of research that already exists — not replace it with opinions or black-box outputs. That’s why Pearl AI is intentionally grounded in the existing evidence base, including peer-reviewed tutoring research, proven program design frameworks, and work from leading academic and research partners.
- Our AI doesn’t just generate answers; it shows its work. Admins can see which research informs a recommendation, how conclusions were formed, and why a particular insight surfaced
Showing your work was non-negotiable in class growing up, and we think it’s non-negotiable with AI because confidence comes from transparency. This ties into a concept we care deeply about: Insight Latency.
- For us, insight latency isn’t just about dashboards. It’s the gap between when tutoring happens and when leaders can act on what they’re seeing.
Historically, that cycle takes months — sometimes six to twelve. Our goal is to compress it to as close to zero as possible, not by skipping rigor, but by embedding analysis and research directly into everyday workflows.
Pillar #3: Give Humans Their Time Back
We’ve talked in the past about AI being a “50× multiplier.” That’s still true, but we believe the real power of AI is giving people their time back.
No one got into tutoring to clean data, normalize files for reporting, chase attendance logs, or schedule across multiple systems.
They got into tutoring to build relationships, support students, and use data to improve outcomes.
Pearl AI is designed to take the most mundane, low-impact, and emotionally draining tasks off people’s plates,so they can focus on the work that really matters.
How This Philosophy Shows Up in What We’re Building
This philosophy isn’t abstract.
It directly shapes what we build, who we build it for, and just as importantly, what our tools can — and cannot — do.
- You can see it clearly in our first AI effort: the AI Admin Agent. It’s why we’ve focused on advice grounded in evidence, insight delivered in minutes rather than months, and tools that unlock human energy — not just surface information.
We’ll go deeper into those ideas and into specific agents in the posts that follow. This one is about the “why.”
This Is Just the Starting Line
AI in education is still early. We don’t pretend to have all the answers, but we are very clear on our principles. This philosophy will remain the reference point for every AI tool we build at Pearl. It’s how we make decisions, set boundaries, and stay honest about what AI should and shouldn’t do.
We’ll keep sharing what we’re learning — openly, honestly, and alongside our partners.
More to come soon.