For schools

The one place a kid uses technology to think harder — not less.

Your teachers are naming it already: kids hand their thinking to a screen, and AI finishes the job before the skill can form. SPARK turns that exact technology into the reason a kid thinks more— because you can't direct a tool you haven't thought through.

It already works: Emeka, age 7, ran the loop and built Star Math — live today, measuring her real growth.

For schools · the program

We run it, start to finish.

An 8-week build program, grades 4–6, one cohort of about 12. We run every session start to finish — zero lift for your staff — and you get a full written cohort report. It comes in through your school’s approval process, with vetted, safety-first AI.

You provide
  • One cohort and a weekly session block
  • A room with internet + student devices (or let’s discuss options)
  • A staff contact for scheduling & consent
We bring (everything else)
  • Both coaches — we run every session, start to finish
  • The full 7-move method + a builder’s workbook per kid
  • The supervised AI build, Demo Day, and your written cohort report
Supports your classroom

What your teachers taught Monday, kids use by Friday.

Expeditions can start from your own curriculum: we ask the kids what they learned this week, then build something real with it. Recalling it is a refresher; building with it is transfer — the lesson lands a second time, a different way, and the subjects connect inside one build.

And it never turns into more homework: the rule is never “practice it again” — always build something with it that helps someone real.

Already proven

Star Math started exactly here — school’s math practice, rebuilt by a 7-year-old into a tool her whole class uses.

The most boring part of the school day became the win the whole room shares. That’s what “supporting the classroom” looks like when a kid does it.

Safety first

How we keep kids safe with AI.

The questions your district will ask — answered before they're asked.

Supervised, always

Two coaches with the cohort, every session — about a 1:6 adult-to-student ratio. The kid directs the tool while a coach guides; never an unsupervised chatbot for a child.

One coach-run workspace

Kids build through a single coach-operated account — no child registers a personal login, so the under-13 question is handled at the source. Every build is reviewed before it goes live.

No student data leaves the building

We don’t collect identifying information, and builds live in the coach workspace — not on kids’ own accounts.

Consent first — and we supply the form

A ready-to-use, opt-in parent/guardian consent form (publishing a build + Demo Day photos), in English & Spanish. Your staff sends it home; you approve the wording first.

Cleared & accountable

The coaches already hold the clearances their current school programs required, and we complete any additional district paperwork before week one.

The deliverable

Your cohort report — success, in writing.

At week eight you hold the thing you can put in front of your superintendent, your board, and next year's funding. Here's exactly what's in it:

  • 1Every student’s finished tool — with a live link
  • 2Each kid’s Skill-Check growth: baseline vs. latest, in their own app’s data
  • 3Attendance & completion for the cohort
  • 4The seven named thinking-moves each child demonstrated
  • 5Demo Day turnout — family & staff
  • 6“Who it helped” — the real people each build served
Scheduling

It fits the week you already have.

See how a session runs →
Lunch & Build
45 min · during school

The flywheel — keeps the build warm and the crew connected mid-week.

After-School Build
90 min · after dismissal

The engine — the deep build block is where real creations get made and Demo Day lives.

A classic 75-minute weekly block works too — we adapt to your master schedule, not the other way around.

Standards & funding

Aligned to the boxes you have to check.

ISTE

Digital-age skills — kids direct technology instead of being directed by it.

CASEL

The empathy core — every build starts with “who’s having a hard time?”

Title IV-A

Defensible for Student Support & Academic Enrichment funding if you scale it later.

Principal FAQ

The questions you're actually going to ask.

Is the AI safe for grades 4–6?

Yes — because it’s never unsupervised. Two coaches run a cohort of about 12 (a 1:6 ratio); the child directs the tool, the coach guides it, and every build is reviewed before it goes live. Kids work through one coach-operated workspace with no personal logins, so the under-13 question is handled at the source.

What happens to student data?

Nothing leaves your building. Kids don’t create personal accounts and we don’t collect identifying data. A child’s build is published only with explicit, opt-in parent/guardian consent — on a form we provide and you approve.

Who actually runs it — does it cost my teachers time?

No. Both coaches run every session start to finish. Your staff’s only job is scheduling and consent. If you later want it to live in the building, we’ll train your teachers to co-run it.

Will this reach every student, including my IEP and ELL students?

Yes, by design. The entry point is a problem from a kid’s own life — no reading level required — so every student can start. Faster builders are pointed at lifting classmates, not racing ahead, and everyone makes something real. We recommend a mixed cohort for exactly this reason.

How is this different from a coding club or an app?

The app is just proof. What lasts is the thinking — a named, repeatable loop a kid runs on a math problem, a friendship, or a free throw long after the 8 weeks end. We teach the brain; the build is where they practice it.

Is the evidence real?

It works, and it’s real: a 7-year-old ran the full process and built Star Math, which measures her own growth on a repeated timed Skill Check (baseline vs. latest, in the app’s own data — see it live at star-math.vercel.app). It’s a focused program with a growing body of proof, and the cohort report documents every round.

Can we start small?

That’s the offer: one cohort, 8 weeks, fully run by us. The cohort report makes the case for the next round.

Let's pick a start window.

A 20-minute walk-through, in person or by phone — we bring everything else, from the workbooks to the supervised build to your cohort report.

Start the conversation →
Coach Lamont Kirton
lamont@communitynyc.org
347-385-9689
Coach Kareem Akabashorun
akabashorunk@gmail.com
347-942-7119

Spark Builders · in partnership with Working Wonders Starting Home