MCAT Learning Lead

Position Overview: The MCAT Learning Lead is the first hire on Premed Catalyst's MCAT program, which we are building from scratch. You own the education side: what students study, in what order, and which methods get them to mastery fastest. You start with our question bank. During the build, the bank is mainly our measurement layer for problems no MCAT tool solves today: mastery per subtopic, error patterns that route students to the right next task, a CARS mistake journal, and confidence calibration. You decide which topics it covers first and the standard every question and explanation is held to, then help build the team of subject-matter experts who write them. You set the research agenda for how our students learn and test it through experiments, using tutor and student feedback to decide what changes next. You work alongside Zach, our CEO, and Karam who works alongside our student success team.

Question Bank & Curriculum: Decide which topics the question bank covers first across all four MCAT sections (CARS, Psych/Soc, Chem/Phys, Bio/Biochem), based on exam frequency, student error patterns, and difficulty. Pick the starting section within your first 30 days, and within 90 days recommend a scaling plan with goals and dates attached: bank size, budget and timeline, production method, staffing, and first section to release. Any learning result the plan relies on must show up in at least two pilot groups or with two tutors. Define what high-quality questions, explanations, and content look like, build a quality rubric that can be applied consistently, and measure every question against it. Within 30 days, the rubric and a standard for explanations, including feedback on each reasoning step of a question, are written and tested on AAMC and competitor questions. Within 60 days, high MCAT scorers pick out 30% or less of our questions in a blind mix with real AAMC questions. Find the best way to teach each MCAT section, including sections that call for a different method.

Learning Experiments: Build and own a research agenda from the student problems no MCAT tool solves today, and decide which ones to test first. Design and run experiments, content-based or student-based, on which methods get students to mastery and improve scores. Target: students gain a median of 3+ more points per AAMC section over 6 weeks than a comparison group you design, such as each student's untreated topics or matched students using only UWorld and AAMC.Run a weekly experiment cycle: a written hypothesis and prediction each Monday, judged on recall 72 hours to 1 week later, reps-to-mastery, and repeat errors within 14 days. Track 80% passage completion and a weekly burnout rating alongside every test. Define what mastery means for each subtopic and test it. Starting proposal: 90%+ on a mastery check, correct on 3 separate days, and still correct at the last scheduled review. Within 30 days, set up the per-answer record every experiment runs on: one row per answered question with student, question, subtopic ID, correct or not, seconds spent, confidence (1 to 5), and date. Use student performance data to identify the specific metrics and actions that predict score improvement, and recommend changes to what and how students study. Within 90 days, show that accuracy on our questions predicts AAMC performance at a correlation of 0.8 or higher, measured on at least 30 students, reported with its confidence interval, and checked against AAMC questions held out from practice. Diagnose mixed results: when five students improve 8+ points and three show no improvement, figure out why and fix it. Document findings: what works, what doesn't, and what's still unknown. Help Us Build the MCAT TeamPlan the subject-matter experts you need, by subject, and help hire them: med students and recent graduates who scored 518+ on the MCAT. Give us the SME criteria within your first 30 days, so the team can write and review the first 20 to 30 questions within 60 days and reach 3,000 questions within 6 months of being staffed. Figure out what makes a great tutor, and build systems to replicate it, including session structure and feedback loops. Work with our development team to inform what gets built: the features students need and the tools tutors use.

Student & Tutor Feedback: Collect feedback from tutors and students, and turn it into changes to the student journey.

Required: A degree in education, learning design, cognitive science, or a related field. Experience working directly with students. Ability to design experiments with small samples and a comparison group, interpret data, and turn findings into strategy. Experience managing people who don't report to you, such as contractors and subject-matter experts. Comfort with ambiguity: you try a new approach before you know it will work, and adapt quickly when the situation changes. Self-directed: you make decisions and solve problems without waiting for direction. You have cut the time or cost of producing something without lowering its quality, and you prefer simple systems built around the few metrics that matter. You have held a quality bar that meant rejecting work, even when it put you behind schedule. You catch errors in technical content before it reaches students. You have planned and delivered a multi-month build against a timeline. PreferredA Master's degree, PhD, or postdoctoral research experience in education, learning design, or cognitive science. Experience with high-performing students. Experience with the MCAT, medical education, or high-stakes test prep. An MCAT score of 515+.Who This Is Not ForAnyone who judges a question by whether it feels right, without a written standard to check it against. Anyone who keeps a study method because it worked for them, after the data shows it isn't working for our students. Anyone who needs a finished curriculum or a defined process handed to them. This program is being built from scratch.

What We Offer: Fully remote Full-time