Skip to content
AI College Prep Academy

ACT/SAT Bootcamp · Inquiries Open

Senior engineer-led · Built on the same AI platform working professionals use · Inquiries open for 2026–2027 remote cohorts

Live expert teachers plus a 24/7 personal AI tutor for every student.

A new generation of ACT and SAT bootcamp. Each student builds their own AI tutor on day one — trained on their weaknesses, available every hour they study, and built to stay with them through retakes, AP exams, and college applications. The AI tutor is included, not sold separately.

High school student studying for the SAT at a wood desk with a laptop, a printed practice test, and a graphing calculator — soft natural light, focused expression

Why this is different

Every other ACT or SAT bootcamp gives your child one teacher for fifteen-plus students. We give them a personal AI tutor on top.

The legacy bootcamp model has not meaningfully changed in twenty years. Bigger classes, recycled question banks, the same week-by-week curriculum no matter which student is sitting in the seat. We rebuilt the format from the inside out, around what AI agents make possible now — and what working engineers actually use AI for every day.

A personal AI tutor that knows your child

Every student builds their own AI tutor during week one of the bootcamp. By week eight, that tutor knows their specific weakness patterns — not generic categories like 'algebra,' but precise patterns like 'consistently misses inference questions in dense reading passages.' Practice questions are generated against those exact gaps. No fixed question bank.

vs. a one-size curriculum no matter which student

24/7 — every hour your child actually studies

Test prep doesn't happen in class. It happens at 10 p.m. on a Tuesday when your child is stuck on a passage they don't understand. The AI tutor answers, explains in the style they learn best, and tracks the question so it shows up again three days later. The tutor lives in your child's AI account and is available from any device with internet.

vs. tutor office hours that end at 9 p.m.

It remembers every mistake so they don't repeat it

Standard test prep treats every practice session as fresh. We use the memory features of modern AI to track every wrong answer over the full eight weeks — and resurface those exact question types at spaced-repetition intervals (three days, one week, two weeks) so the patterns actually stick.

vs. drill packets the student forgets by Friday

The AI tutor stays with them after the bootcamp ends

When the eight weeks finish, the tutor is still in your child's AI account. They keep training it for ACT or SAT retakes, AP exams, college essay drafts, future test prep — through college applications and beyond. You are paying for a study system that compounds for years, not a packet that gets recycled.

vs. eight weeks of value, gone when the bootcamp ends

How the AI tutor works

Three things the AI tutor does that fixed-question test prep cannot.

We build each student's AI tutor on the same AI platform working professionals use to ship production software. The tutor is not a chatbot, and it is not ChatGPT. It is a long-running AI agent with memory, with access to your child's specific test-prep history, and built to generate new material against their exact weaknesses.

Knows exactly what your child struggles with

Every wrong answer your child gives — across every diagnostic, quiz, and full-length practice test — feeds into the AI tutor's memory. After two weeks the tutor has identified granular patterns no human tutor would catch this fast: compound-interest questions with quarterly compounding, inference questions in dense historical passages, function-graph transformations in reverse.

Built on AI memory and project features — the tutor accumulates a longitudinal model of your child specifically.

Practice questions made just for your child

Once the tutor knows the patterns your child keeps missing, it generates targeted practice questions matched to those exact patterns — not random drills from a fixed bank. Your child builds a personalized, effectively infinite practice library that attacks the specific gaps they keep falling into.

Generates new questions every session based on weakness patterns, calibrated to the question style of the Digital SAT and current ACT.

Teaches the way your child learns best

Each student trains the AI tutor on how they learn. Visual learners get diagram-heavy explanations and worked examples. Step-by-step learners get logical derivations. Conceptual learners get story-form explanations tying ideas to the real world. Over eight weeks the tutor gets better and better at explaining the way your child specifically processes information best.

The student is in charge of shaping how the tutor teaches — a meta-skill that pays off long after the SAT or ACT.

The 8-week arc

Four phases that build the test-taker — and the AI tutor — together.

Most test prep starts at 'here are the question types.' We start earlier: structure, baseline diagnostics, and your child's AI tutor account. By the end of week eight, your child has both the score-prep work AND a personalized AI study agent they keep using for years.

    Weeks 1–2

    Foundation

    Test structure deep-dive (Digital SAT module timing or ACT section structure), a full-length baseline diagnostic, and the setup of each student’s personal AI tutor account. By the end of week two, your child has a tutor they can talk to.

    • Baseline diagnostic on an official-format full-length practice test
    • Personal AI tutor account set up and configured
    • Test-day mechanics: pacing, registration, allowed materials, accommodations process
    • Initial weakness areas documented in the AI tutor’s memory
    Weeks 3–4

    Pattern Recognition

    Granular weakness analysis. Instructors teach the question-type taxonomy in plain English; students teach their AI tutor to recognize and explain those patterns back to them. Practice volume increases.

    • Each student names their five biggest weakness patterns explicitly
    • AI tutor is trained to spot those patterns in a new question and walk through them
    • First mid-bootcamp full-length practice test under timed conditions
    • Spaced-repetition resurfacing of weeks 1–2 misses begins
    Weeks 5–6

    Deepening

    The AI tutor generates progressively harder targeted practice against each student's specific gaps. Students teach the AI tutor to explain in their learning style — visual, step-by-step, or conceptual — which makes the tutor a better teacher week after week.

    • Custom practice problem sets generated by the AI tutor every session
    • Students refine how the tutor explains: visual diagrams vs. derivations vs. story form
    • Section-specific test strategy: pacing rules, when to skip, when to guess strategically
    • Second mid-bootcamp full-length practice test — measurable progress on prior weaknesses
    Weeks 7–8

    Mastery

    Advanced AI tutor features: cross-session memory of every mistake over the full eight weeks, custom rubrics for writing prompts, and personalized test-day playbooks. The final two weeks are about confidence, consistency, and a polished plan for test day.

    • Final full-length practice test in official conditions
    • Personal test-day playbook generated by the AI tutor
    • Plan for ongoing tutor use after the bootcamp ends — for retakes, AP exams, college apps
    • Parent showcase: students walk parents through what their AI tutor can now do

Choose your track

Parallel SAT and ACT cohorts. Same AI tutor framework. Different curriculum.

Families pick which test their child will focus on — or both, for students keeping their options open. The instruction is calibrated to each test's specific structure (the Digital SAT and the current ACT), but the AI tutor architecture, learning-style adaptation, and 24/7 support model are identical across tracks.

Digital SAT

Built around the current Digital SAT format — adaptive Reading & Writing module, Math module with built-in calculator, and the section-timing realities of the new test.

The Digital SAT changed how students should prepare. Adaptive module difficulty means second-module performance depends on first-module accuracy. Embedded calculators change which math shortcuts matter. We teach the test that students will actually sit for.

  • Adaptive-module strategy: how first-module pacing shapes second-module difficulty
  • Reading & Writing question taxonomy with AI-tutor pattern memory
  • Math section: when to use the embedded calculator, when not to
  • Section-by-section pacing built around real test-day conditions

ACT

Built around the four ACT sections (English, Math, Reading, Science) plus optional Writing, with strategy calibrated to the ACT’s distinct timing pressure and question style.

The ACT rewards a different set of test-taking habits than the SAT. The Science section is really a data-interpretation section. The English section punishes overthinking. The ACT timing pressure is more aggressive. We teach the test on its own terms — not as 'the SAT with science.'

  • Section-by-section pacing for ACT’s tighter time limits
  • Science section taught as data interpretation, not biology recall
  • English section: grammar patterns and rhetorical-skill question types
  • Optional Writing section coaching for students who need a writing score

SAT and ACT

For students taking both tests, or keeping their options open during junior year — the AI tutor framework supports both tracks in parallel with full-length practice tests for each.

Some students benefit from preparing for both — different colleges weight them differently, and some students naturally perform better on one. The AI tutor handles both test architectures in parallel, with separate weakness-pattern memory per test so progress on one doesn't get tangled up with the other.

  • Parallel curriculum modules for both tests
  • Separate AI tutor memory streams per test — no cross-test confusion
  • Comparative diagnostics to surface which test plays to your child's strengths
  • Test selection conversation in week three with full-length-test data in hand

Format and delivery

Eight weeks of intensive small-group instruction with a personal AI tutor running in the background.

An eight-week summer intensive — Monday through Friday — built so families get serious volume (real instructional hours, real practice tests, real strategy reps) without the per-hour cost of one-on-one tutoring. The AI tutor extends every classroom hour into the rest of the week.

Daily structure

Each weekday is a structured arc: subject-block instruction in the morning, working lunch, AI-tutor practice session in the early afternoon, and test strategy or full-length-test segments to close. Same expert instructors, same students, same AI tutor framework all eight weeks.

  • Two morning subject blocks — Math, plus Reading & Writing (SAT) or English + Reading + Science (ACT)
  • Working lunch — students bring their own
  • Personal AI tutor practice session — every student, every day
  • Afternoon: test strategy, mock test segments, individual coaching

Class composition

Small instructional groups so every student is known by name and progress pattern. We hit the ratios test prep families actually want — not the lecture-hall class sizes that legacy bootcamps default to.

  • Small instructional groups — each cohort built so students get real instructor attention
  • Multiple specialist instructors per cohort (math + verbal + test strategy)
  • Teaching assistants for AI tutor setup and live troubleshooting
  • Parent showcase in week eight — students demo their AI tutor to families
Small group of high school students at a long study table with laptops and printed practice tests, an instructor working with one student’s AI tutor on screen — bright daylit classroom

What students leave with

A study system that compounds — not a packet that expires.

Every student finishes the bootcamp with the same set of deliverables. The AI tutor is the centerpiece — the asset your child keeps using long after the eight weeks end.

A personal AI tutor they own

A configured Claude project containing their full diagnostic history, weakness patterns, generated practice library, and personalized explanations. Lives in their AI account. Stays with them.

200+ hours of small-group expert instruction

Real instructor hours with specialist teachers across math, reading & writing (SAT) or English + reading + science (ACT), and test strategy — a higher instructional volume than the legacy bootcamp baseline.

Multiple full-length official-format practice tests

Diagnostic, mid-bootcamp, and final practice tests — every full-length test run under official conditions and analyzed by the AI tutor against each student’s specific weakness patterns.

The skill to keep training the AI tutor

After the bootcamp ends, your child knows how to keep training the AI tutor — for ACT or SAT retakes, AP exams, college essay drafts, future test prep. The eight weeks are the on-ramp; the AI tutor is the system.

Timeline and availability

Pilot cohorts now. In-person bootcamps coming in summer 2028.

We are intentionally piloting the AI-tutor curriculum remotely first, with smaller cohorts of families who want to be part of the early generation. In-person bootcamps are planned for California and Illinois in summer 2028 — and inquiry waitlists are open for both.

    Summer 2026 · Fall 2026

    Remote pilot cohorts

    Small remote cohorts running on the same eight-week format. We use these cohorts to validate the curriculum, refine the AI tutor setup, and collect score-improvement data for families who want a paper trail before in-person launches.

    Inquiries open
    Summer 2027

    Expanded remote cohorts

    Wider remote enrollment as the curriculum solidifies — open to families anywhere in the U.S. with reliable internet. Same instructors, same AI tutor framework, same eight-week pedagogical arc as the in-person 2028 launch.

    Inquiries open
    Summer 2028

    First in-person bootcamps · California and Illinois

    The inaugural in-person bootcamps run in California and Illinois. Families who want their child in the first in-person cohort should inquire early — seats are intentionally limited so the small-group instructional ratios hold.

    Waitlist open

Who is teaching this

Built by an engineer who ships production AI for a living.

Jinwoong Lee — a 2018 UC Berkeley graduate — is a senior software engineer with eight years of experience building production software at companies that include Amazon and Ring. He also founded CareMAR, software that helps small care homes and hospice facilities replace paper-based medication tracking with an AI-assisted app — built for owners who aren’t technical, in a regulated healthcare environment. That combination — production AI engineering plus a track record of teaching non-technical users to actually adopt AI — is rare in test prep, and it’s the foundation the AI-tutor curriculum is built on.

8 years of production software engineering

Companies that include Amazon and Ring. Currently a senior software engineer. The same engineer who builds AI systems at work designs the AI tutor architecture students train in the bootcamp.

Ships AI to non-technical users in a regulated industry

Founder of CareMAR (caremar.us), AI-assisted software for small care homes and hospice facilities operating under strict healthcare-privacy rules. Building AI that non-engineers actually use is exactly what the bootcamp teaches students to do.

11 years of private education practice · 80+ students placed

Eleven years coaching students from his own private practice; mentees placed at Harvard, Stanford, Princeton, Columbia, Brown, Dartmouth, Cornell, UC Berkeley, UCLA, Northwestern, Georgia Tech, and other top-twenty universities. The teaching style in the bootcamp is the same one those placements were built on.

The combination is the point

Most test prep founders are educators who have never built AI. Most AI engineers have never taught a teenager. The ACT/SAT Bootcamp curriculum exists because one person sits at both tables — and that combination is what makes a working personal AI tutor for every student possible.

Portrait of Jinwoong Lee, founder of AI College Prep Academy and CareMAR — senior software engineer with eight years of production software experience.

FAQ

Common questions from families.

Inquiries open · Remote pilot 2026–2027 · In-person CA + IL summer 2028

Be in the first generation of students with a real personal AI tutor for the SAT and ACT.

If you want your child in one of the remote pilot cohorts in 2026 or 2027, or on the waitlist for the inaugural in-person bootcamps in California or Illinois in summer 2028, start with a short inquiry. We get back to families within two business days.