Weekend AI Seminars · Open Enrollment
Senior-engineer-led AI seminars for private school communities. We teach the Claude platform — Skills (saved AI recipes anyone can author), Cowork (shared workspaces where Claude is in the room), Projects (your textbook + notes uploaded once, grounded answers from YOUR sources). The systems your school actually keeps. Not another prompt course.
Communities · Facebook
4 groups for parents, teachers, college-prep families, and school leaders. 293+ already ahead of you.
Why this seminar, not another one
There are a lot of "AI consultants" in your inbox. Most are ex-marketers, ex-coaches, or ex-recruiters who took a six-week prompt course. The seminars are different in four ways that compound.
Built production AI systems used by tens of millions of customers at Amazon, Ring, and adjacent product orgs. Knows what AI actually ships, what fails in real users' hands, and what's hype. Engineers who have shipped AI in front of tens of millions of users teach AI fundamentally differently than someone who finished a six-week prompt course.
vs. coach / recruiter / marketer-turned-AI-consultant
Chat sessions evaporate. Workflows compound. We teach you to build saved systems that run every Friday — gradebook → personalized parent emails, lesson plan + roster → differentiated versions, calendar + homework load → weekly study plan. Real infrastructure, not prompts you forget by Monday.
vs. "here are 50 ChatGPT prompts to copy"
We teach the Claude engine — Skills (think saved AI recipes anyone in your school can write), Cowork (think a Google Doc with Claude already in the room), Projects (think a digital filing cabinet Claude has read). Then we wire in supporting tools where they earn their place — Gemini when the data lives in Google Workspace, Perplexity for grounded internet research, others as the supporting cast. Not as equal-billing chatbots.
vs. "AI is just a chatbot you talk to"
The rare combo: builds AI for a living, AND has spent eleven years teaching technical concepts to teachers, students, and parents. Most production engineers cannot teach. Most teachers haven't shipped AI. We have placed 80+ students at top universities using exactly the teaching style we run in these seminars.
vs. brilliant engineers who can't explain it / generic teachers who don't ship
The Claude-native engine
We teach the Claude platform because it's purpose-built for how schools actually work. The three pieces below are the engine — and we describe each in plain English so you can decide whether it fits before you walk into the room.
Anthropic
Best for
Saved AI capabilities · Build once · Share with everyone
Think of a Skill like a saved recipe Claude follows whenever you ask. Write 'how to grade an essay using my rubric' once, save it as a Skill, and every teacher in your school can use the same Skill the same way every Friday. Build a 'study buddy for AP Bio' Skill — every student in the cohort invokes it. No coding required: anyone can write a Skill in plain English the same way you'd write instructions for a new substitute teacher. This is how AI starts compounding across an institution instead of evaporating in one chat tab.
Anthropic
Best for
Shared AI workspaces · Multiple people + Claude · Real time
Think of Cowork like a Google Doc with Claude already in the room. A teacher and the entire class see the same conversation and contribute together. A family — parents, child, and Claude — work on the college list at the same time. A study group of three students plus Claude tackles the same lab report together. Cowork is what turns Claude from a chat tab one person uses into an actual team member, with shared memory across sessions.
Anthropic
Best for
Your real sources · Grounded answers · No more hallucination
Think of a Project like a digital filing cabinet you upload your real materials into. Your AP Bio textbook, your class notes, your school's website, your family's college list — drop them in once. Claude reads everything in the cabinet before answering, so its answers stay grounded in YOUR sources, not the open internet. The textbook lives inside the AI; you stop pasting things in every time. Every answer points to a specific page in your uploads — verifiable, not made up.
The supporting cast we wire in
Real workflows chain these into Claude / ChatGPT / Gemini — that’s the multiplier.
Gemini
When the data already lives in Google Workspace — Gemini's already inside Docs, Drive, and Gmail.
Perplexity
For grounded internet research with real citations — pairs with Claude for synthesis.
NotebookLM
When you need cross-document Q&A across many sources — a different shape of grounded knowledge base.
ChatGPT
For voice-mode practice and image generation when those specific features matter.
Cursor
For long-form writing where Claude lives inside a real editor with you.
Zapier / Make
The connective tissue — wire any of the above into your gradebook, calendar, or email.
AI agents you actually walk out with
These are not screenshots from a deck. They are the actual systems we build in the seminar — saved on your laptop or in your AI account, running on your data, after you leave. Each one is what an experienced production engineer would design — chained tools, source-grounded outputs, guardrails for when AI is wrong.
Practice mock college interviews with a voice AI that asks real questions and grades you on the answers.
A voice-mode AI agent that runs a thirty-minute mock interview in the style of Harvard, Wharton, Stanford, or any school you target. It asks the questions admissions interviewers actually ask in 2028, listens to your spoken answers, transcribes them, and grades you on framing, specificity, and the parts you skip. Plays back the recording with timestamped feedback. Adaptive — gets harder on the topics where you're already strong, drills harder on the ones where you stumble.
What it does
A test-prep agent that learns how YOU learn and re-asks the questions you forget at the right intervals.
Reads your prior practice-test results, identifies the topic patterns where you lose points, and builds custom problem sets that target your specific gaps. Tracks every wrong answer and re-asks it at spaced-repetition intervals — three days, one week, two weeks — so the questions stick. Detects which explanation style works best for your brain (visual, step-by-step, intuitive) and defaults to that style. Generates a weekly progress report that shows your improvement curve and what to focus on next.
What it does
An AI that has actually READ your textbooks, your class notes, your handouts — and answers ONLY from your sources.
A Project (in Claude or NotebookLM) that has read every chapter of your AP Bio textbook, every set of class notes, every handout your teacher uploaded — and answers your questions GROUNDED in those sources. Returns answers WITH page numbers and excerpt quotes, so you can verify nothing was hallucinated. Generates flashcards with source citations. Creates practice tests at increasing difficulty. The textbook lives inside the AI; you stop pasting things in every time.
What it does
Pulls a weekly list of summer programs, competitions, and ECs that match your child's profile and your family's constraints.
Reads your child's current profile (activities, academic interests, college goals), then surfaces summer programs, competitions, and extracurricular opportunities that match. Filters by application deadline, fit, cost ceiling, travel willingness, and the parent commitments you actually have. Ranks the list by fit + admissions impact. Sends you a weekly digest of new opportunities with deadlines and applications links. The kind of college-counselor work that costs hundreds an hour, running automatically every Sunday.
What it does
Walks your child through a structured interview, then drafts their first college resume in three formats.
Runs a structured intake interview with your child — the activities, projects, leadership, hours, awards, real impact. Drafts the resume in three formats: Common App activities list, LinkedIn-ready professional resume, and a Harvard-style narrative resume. Surfaces gaps in the profile (e.g., 'you have three tech ECs and zero community service — here are options to address it before junior year'). Generates supplementary essay prompts based on the resume content. The output is something an experienced college counselor would charge for hours of work to draft.
What it does
Builds a reach/match/safety table, surfaces demonstrated-interest signals, and runs a 12-month application timeline.
Reads target school websites and admissions data, builds a personalized reach/match/safety table from your child's profile, and surfaces the demonstrated-interest signals each school weights (visits, supplemental engagement, alumni interviews, EA vs. RD timing). Builds a 12-month application timeline with milestones — testing windows, supplement themes, recommendation requests, financial-aid forms. Tracks which schools have which essay prompts and groups them by theme so your child can write essays once and reuse the angles.
What it does
A Claude Skill that maps APs and electives to your child's college targets, workload tolerance, and academic trajectory.
Reads your child's transcript, current course load, and target college list. Outputs a 3-year course plan that balances rigor (per-college expectations), workload (without burnout), and depth (where your child shows real interest). Updates each semester as preferences shift. The kind of strategy work a top-tier college counselor delivers — running as a Skill any parent in your school can install.
What it does
A Claude Skill that pairs you with the right teachers and drafts the brag sheet they need to write a strong letter.
Analyzes which of your teachers know you best in which contexts, identifies the strongest two-teacher pairing for your application strategy (typically humanities + STEM), and drafts the brag sheet — concrete projects, anecdotes, evolution-of-thinking moments — that turns a generic recommendation into a great one. Plus the thank-you note draft you actually want to send afterwards.
What it does
A shared Claude Cowork space the whole family lives in — weekly Sunday discussion prompts based on what your kids are actually using.
A multi-user Claude Cowork workspace your family shares. Generates a weekly Sunday-night discussion prompt based on the AI tools your kids report using that week. Surfaces patterns to discuss (when to verify, when to use AI, when to put it down). Keeps a running family AI-literacy notebook that compounds — by junior year your child has a documented track record of thoughtful AI use that admissions readers can verify.
What it does
Yes, you actually build these. The afternoon hands-on workshop is where these go from slides to running on your laptop. By the time you leave, the agent runs on your data, on your account, indefinitely.
What you walk out building
Every session addresses teachers, students, and parents in the same room — because the most useful conversation about AI happens when all three sit together. Each leaves with saved workflows running on their own laptop, not prompts they’ll forget by Monday.
Build the systems that compound. Not prompts you copy weekly.
Stop pasting your gradebook into ChatGPT every Friday. Build the saved workflow once, then run it every Friday in three minutes — with parent emails written in your voice, drawn from real grade data, and edited in one pass.
Concrete builds you walk out with
Build a study system. Not a homework cheat path.
AI as a thinking partner, scoped to YOUR sources — not the open internet. Build a study assistant that has actually read your AP Bio textbook, a research workflow that finds and reads five academic papers, and a college-research system that surfaces signals other applicants miss.
Concrete builds you walk out with
Be the AI-fluent parent. Build the systems your family runs on.
Be the parent who knows what their child uses AI for, AND who uses AI better than them. Build college-research systems, family logistics workflows, and a homework-help assistant that asks your child clarifying questions instead of giving them answers.
Concrete builds you walk out with
Community · Facebook
Quick weekly tips, links to new Claude Skills, parent Q&A threads, and the kind of conversations you don't get on a one-time seminar day. New groups open as the community grows — if a group isn't live yet, sign up for early access and you'll be invited the day it opens.
Parents · AI Literacy
108 parents on the early-access list
Weekly AI tips that actually save you time — school emails, college research, family scheduling. Light Q&A every Friday. Curated by a senior engineer (not an AI consultant).
Join the waitlistTeachers · Claude Skills
64 teachers waitlisted
A working group for teachers authoring reusable Claude Skills — grading rubrics, lesson scaffolds, parent-comm drafts. Skill of the week, peer review, no vendor pitch.
Join the waitlistCollege Prep · ACT/SAT
79 parents waitlisted
Honest conversations about using AI to support test prep, find tutors, and prep college essays — without crossing into cheating. Senior-engineer-led, parent-readable.
Join the waitlistSchool Leaders · Policy
42 leaders waitlisted
How to think about AI policy, acceptable-use, faculty PD, and parent communication — from people who actually ship AI for a living. Quarterly long-form, weekly short notes.
Join the waitlistWe won't use your email for anything except the group launch invite. Promise.
A Saturday at our seminar
Every session is structured the same way — a live morning framework session for the room, followed by smaller hands-on workshops in the afternoon where you actually build the workflows on your own laptop.
Senior-engineer-led demos. Real prompts, real outputs, real workflows you can start using on Monday. Audience-aware — teachers see teaching workflows, students see student workflows, parents see family workflows. Multi-tool from the first demo.
Smaller groups, your laptop, our facilitator. You leave with workflows actually running on your own setup, not theory in a notebook. Take-home prompt + Project library you keep.
An engineer's perspective
An engineer who ships production AI teaches AI differently than a consultant who took a course. These are the principles we built the curriculum around.
We test AI outputs daily — in production, with real users, where mistakes have real costs. The seminars teach you the failure modes most consultants don't know about: hallucinations on numbers, citation fabrication, prompt injection, and the way models silently change behavior between releases. Knowing when AI is wrong is half the fluency.
Production AI systems route between five to ten tools — language models, retrieval systems, code execution, structured outputs, downstream APIs. The same pattern works for school workflows: Claude reads the textbook, ChatGPT generates the quiz, Gemini drops it into Google Classroom, Zapier emails the parents. We teach the chain, not just the chatbot.
An impressive AI demo that takes longer than the manual task is a net negative. We optimize for what actually compresses your time and attention — short, well-scoped workflows you run ten times a week, not flashy multi-step pipelines you'll abandon after the second try.
How we teach AI
Every audience leaves the seminar understanding the difference. We don't teach shortcuts that erode skill — we teach workflows that compound real expertise.
Students learn AI as a research and thinking partner. The work stays theirs; the AI accelerates the process. Teachers and parents learn to spot the difference in practice.
Every workflow we teach includes verification — citations, source-checking, fact-grounding. Students leave less likely to trust an AI hallucination than they were before.
We never use real student work, real medical data, or real family information in demos. Your child learns AI safety as a discipline, not as a footnote.
Parents leave knowing exactly what their child is learning, what tools they're using, and how to talk with them about AI's role in their education.
For private school administrators
We co-host weekend AI seminars at private schools across the region. You provide a classroom and a parent email blast; we provide curriculum, instruction, materials, and follow-up resources. Our partner schools use these sessions as part of their parent education calendar — and as faculty professional development.
FAQ
Open enrollment · By season
Whether you're a parent who wants to attend, a teacher who wants to host one, or a school administrator who wants the full series for your community — start the conversation and we'll take it from there.