01
Learn
We build a Learning Portrait from conversation, sample work, parent context, and careful observation.
Artifact: portrait, pacing map, misconceptions, motivating analogies
Baseline diagnostic
Pattern reading
Family context
Personalized Education, Handcrafted
A tutor who learns your child, then teaches like they mean it.
We build a Learning Portrait that captures what actually clicks for your student, then use it to deliver human-led instruction that feels personal and provably effective.
Invitation-basedHuman-ledParent-controlled data
Learning Portrait
Max, age 14
Analogies That Land
Thinks best through sports momentum, scoreboards, and “show me the tape” examples.
Current Friction
Rushes multi-step math when confidence dips; benefits from slower verbal check-ins before writing.
Signals We Track
Week 04Recall after 24 hoursImproving
Error type repetitionShrinking
Independent startsUp 3 of 5 sessions
Next Move
Use one worked example, then hand the board back. End with an off-screen prompt: explain tonight's homework to Dad without notes.
The Promise
What parents want. What is broken. Our answer.
Parents want more than better grades. They want a child who feels increasingly capable, increasingly independent, and understood well enough that learning stops feeling generic.
What they usually get is the opposite: one-size-fits-all instruction, shallow personalization, and software that confuses more screen time with more support.
Our answer is a handcrafted learning system. We build a Learning Portrait, teach through what actually lands, and show the evidence clearly enough that parents can inspect it for themselves.
How It Works
Learn. Teach. Prove.
Three disciplined stages. Each one leaves behind visible artifacts.
02
Teach
Instruction adapts to the student rather than forcing the student to adapt to the system.
Artifact: session notes, off-screen prompts, tailored explanations
Live tutoring
Substrate-guided explanations
Human judgment first
03
Prove
Parents see what changed, what still needs work, and why the next step is the right one.
Artifact: progress letter, work portfolio, reassessment evidence
Leading indicators
Before/after work
Visible outcomes
What Families Get
A premium service menu, not a black box.
Learning Portrait
A living profile of interests, misconceptions, pacing, confidence, and the analogies that actually unlock understanding.
Weekly Progress Letter
A concise parent update on what moved, what stalled, and what we are watching next.
Work Portfolio
A clean record of artifacts, revisions, and problem-solving snapshots instead of vague reassurance.
Off-screen Prompts
When the right next step is a walk, a notebook, a conversation, or rest, the system says so.
Evidence, Not Vibes
Measurement that parents can inspect.
Baseline first. Leading indicators second. Outcomes after that. No theater.
Baseline
We start with a diagnostic, sample work, and a portrait of how the student currently thinks.
Leading Indicators
We watch recall, error types, effort consistency, reading depth, and whether confusion is shrinking in the right places.
Outcomes
Parents see before-and-after clarity, stronger work, and better transfer instead of vague claims about engagement.
“We don't hide the ball.”
Technology With Restraint
Programmed to make people more human.
A few days after his best friend died, the system told its creator to stop working and go write the eulogy instead. That is the design principle in one sentence: if the most human next step is away from the screen, the system should say so.
Safety & Privacy
High-trust by design.
Parent-controlled data and consent at every step
No ads, no resale, no hidden data marketplace logic
Clear AI boundaries: tool for learning, not relationship replacement
Access limited to what is needed for the student and family
Bridge
Same substrate. Different domain.
We built this for the most demanding environment we know: helping a young person learn while parents watch closely. The same discipline now powers consulting work for organizations that need AI grounded in real context, real judgment, and real proof.
Substrate Consulting
AI systems that know the room.
Narrative Environments
Systems that understand the story of an organization well enough to answer in context, not just retrieve documents.
Context Layers
Operational memory that separates signal from noise so the right facts appear at the right moment.
Evidence Loops
Workflows that show whether the system is actually improving outcomes instead of merely sounding intelligent.
Case vignette
Law firm operating memory
For one law firm, we built an internal narrative environment that organized matters, client history, operating nuance, and drafting context tightly enough that it replaced the practical memory burden previously carried by a senior employee.
Talk Substrate ConsultingAbout
Why I built this.
I built Gad Valley Book Club because the people I love deserved something better than generic software and generic tutoring. The work began with family, but the principle now extends further: build systems with enough context to help well, enough restraint to stay human, and enough evidence to earn trust.
Founder: Jud Soderborg
Contact
Two paths. Same standard.
We reply within 2 business days.