ReGenesis
Sign in
The Epistemological Substrate

The Wisdom Corpus.
Humanity's deepest insights,
structured for everyday life.

More than eighteen thousand curated works across nineteen domains of human knowing. Tagged, related, and made retrievable by a developmental intelligence that knows which insight will land for this person, in this moment.

18,000+
Curated works
19
Domains of knowing
6,400+
Wisdom nodes
45
Metadata columns per node
01 — Why it exists

Foundation models read the internet.
That isn't wisdom.

Claude, GPT and Gemini have read everything — but they hold Marcus Aurelius and a Reddit comment with roughly the same weight. Breadth without curation, relationship or developmental awareness is memory, not understanding.

ReGenesis is a developmental intelligence platform, and Sasha — its 24/7 AI companion — is what users actually meet. For Sasha to serve a person's actual growth, in a 3 a.m. identity question, a Monday morning 1:1, a difficult conversation with a teenager, a client decision under pressure, it needs something most AI systems don't have: a curated, structured, developmentally-tagged body of human wisdom that knows what's appropriate for whom, at what stage, through which expression.

Perspective

The ancient traditions, the cross-cultural insights, the centuries of careful human thought — they ground us. The Wisdom Corpus doesn't retrieve information. It retrieves perspective. And structured, tagged, developmentally-aware perspective is something we can actually engineer.

Wisdom lives at every level of a life.

The same disconnection from self, others and the world that creates individual suffering also creates organizational dysfunction and civilizational crisis. These are not separate problems. They are the same crisis at different scales.

A platform that genuinely serves human development has to meet people where they actually live: in the quiet existential question at 3 a.m., and in the ten-minute 1:1 before lunch. In a senior partner’s client strategy, and in a business analyst’s first difficult conversation. The Wisdom Corpus is built to do both.

02 — What it is

Not a database of quotes. A living epistemological substrate.

The Wisdom Corpus is more than 18,000 curated works — books, papers, sutras, poems, talks, practices, traditions — held in a structured knowledge graph. Each work is tagged across 45 dimensions: bibliographic, developmental, epistemological, ontological, relational, application. Each is locatable by who it is for, what transformation it supports, which developmental stage receives it well, and which it should be withheld from — because a teaching that liberates someone at one stage can destabilize, overwhelm, or be spiritually bypassed by someone at another. Premature or misattuned wisdom does real harm. The corpus knows the difference.

The 19 domains of human knowing

Click any domain to see sub-fields · click a sub-field for representative thinkers + traditions

What counts as wisdom

Every entry in the corpus is tested against an Epistemological Compass — a discipline for distinguishing genuine wisdom from content. A wisdom node is an irreducible, convergent, leverage-bearing recognition about the nature of reality — and about what it means to be a human being participating in that reality, co-arising with it, in relationship to it.

If it isn't convergent across independent traditions, if it isn't adversity-tested, if it doesn't produce disproportionate movement when applied — it doesn't make the cut.

Seven tests for a wisdom node
  • IrreducibleCannot be simplified without losing meaning
  • Non-obviousNot what common sense would generate
  • ConvergentFound independently across traditions
  • Leverage-bearingSmall recognition, large movement
  • Adversity-testedHeld by people facing the worst, still found useful
  • Humble about expressionFinger pointing at the moon, not the moon
  • BoundedClear edges about where it applies
The expression hierarchy

One truth. Eleven ways it can be given.

A wisdom node is not a sentence. It is a recognition that can be expressed at eleven distinct levels — each with its own portability, its own receiver requirement, and its own fitness for a moment. The same node — control and aliveness are inversely proportional — appears as a Taoist symbol, a ritual of surrender, the Prodigal Son parable, a Rumi poem, an aphorism ("let go or be dragged"), the practice of meditation, the principle of autonomy support in Self-Determination Theory, the best practice of delegating by outcome, and the specific technique: "What result do we need by when?" Same mountain, different altitudes.

0
Embodied / Pre-representational
Direct transmission through presence and participation. Cannot be fully textualized. The teacher's silence. The held gaze.
1
Symbol / Sacred Image
Compressed meaning in a single percept. Mandalas, yantras, the cross. Bypasses discursive mind.
2
Ritual / Ceremony
Symbol made temporal and participatory. You move through it. Initiation, baptism, vision fast.
3
Story / Myth / Parable
First fully linguistic expression. The mind's native format. Prodigal Son, Coyote tales, the epics.
4
Song / Poem
Language made rhythmic and melodic. Reaches parts of mind prose cannot. Psalms, Rumi, Mary Oliver.
5
Sutra / Compressed teaching
Ultra-portable. High receiver requirement. Memorized and unpacked over years. Yoga Sūtras, I Ching.
6
Aphorism / Koan / Proverb
Maximum semantic density in minimal form. "The Tao that can be named is not eternal Tao."
7
Practice / Discipline
Sustained repeated activity producing embodied realization. Meditation, prayer, training, therapy.
8
Principle
First fully propositional statement. "Autonomy is a prerequisite for intrinsic motivation."
9
Best practice
Domain-specific application. "Delegate by specifying outcome, not method."
10
Technique
Most concrete, discrete, immediately executable. "Ask: what result do we need by when?"
The hierarchy of knowledge

Four levels from the deepest, most compressed truth to its most actionable form. Most enterprise content lives at level 3. The corpus holds all four.

Wisdom node
Irreducible recognized truth. "Control and aliveness are inversely proportional."
Principle
How the node applies in human life. "Genuine motivation requires both autonomy and felt significance."
Expression / Best practice
Domain-specific application in any of the 11 forms.
Technique
Discrete, immediately executable action.
Why this matters at runtime

Sasha doesn't just retrieve "the right wisdom node." It retrieves the node and the right expression level for this person, in this moment. A business analyst in the 45 minutes before a hard conversation needs the technique (level 10). The same person six months later, sitting with a shift in how they hold authority, may need the parable (level 3). The corpus makes that choice structurally, not by accident.

The bestseller on management found the technique. The corpus holds the node. These are not competing — they are different altitudes of the same mountain.

03 — The dual-space architecture

Most platforms work in one space.
Sasha works in both.

Every wisdom node in the corpus carries dual-space coordinates. This is the deepest structural distinction in the system — and the one most coaching and learning products miss. It is the difference between retrieving content and supporting an actual human becoming.

O

Ontological Space

what things ARE

A three-layer model of who this person actually is, right now, at different levels of inertia.

Layer 1 — Constitutional
highest inertia
Nervous system baseline, attachment style, developmental stage (Kegan, Spiral Dynamics, Cook-Greuter held together), core safety orientation, ultimate concern.
Layer 2 — Pattern / State
moderate inertia
Dominant narrative, recurrent thought patterns, habitual emotional loops, defensive structures, relational patterns recreated across relationships.
Layer 3 — Present-moment
lowest inertia
Autonomic state, somatic readiness, emotional weather. Whether this person can receive anything right now — determines if any offering can land.
The layers relate circularly, not hierarchically. A dysregulated nervous system produces a worldview of unsafety, which produces defensive behaviors, which confirm the worldview. Closed loops. Different entry points have different leverage depending on configuration.
T

Transformation Space

what is BECOMING

Ten distinct, independently verified dynamics by which humans genuinely change. Each has its own supporting conditions, contraindications and resource requirements.

  1. 1.Threshold / Initiationvan Gennep, Turner, hero's journey
  2. 2.Subject–Object ShiftKegan, Buddhist mindfulness
  3. 3.Somatic CompletionLevine, Ogden, EMDR
  4. 4.Meaning Dissolution & ReframeFrankl, Kuhn-for-the-self
  5. 5.Relational WitnessRogers, Buber, Ubuntu
  6. 6.Relational RepairGottman, attachment repair
  7. 7.Contemplative Insightdirect recognition — discontinuous
  8. 8.Practice / HabituationAristotle, nervous-system rewiring
  9. 9.Ecological ReconnectionRoszak, Macy, Plotkin
  10. 10.Creative Expressiontransformation through making
Most platforms work inside two of these ten — practice/habituation and relational witness. Sasha works across all ten, and knows which are possible given a person's ontological ground today.
Relational Fit

The Interval Model

When Sasha offers wisdom, it does not ask what does this person need to know? It asks what interval between what they already hold and what we offer will produce the intended movement? Four modes, dynamically weighted by somatic state, relational trust and present-moment readiness:

Confirmatory
unison — feeling seen
Expansive
octave — same truth, higher altitude
Complementary
different tradition, natural harmonic
Productive Dissonance
maximum tension — the koan for the fundamentalist
04 — How it's built

Five intelligence layers on a medallion data pipeline.

The corpus is not an afterthought bolted onto a chatbot. It is engineered as the substrate. Below the surface experience, five layers of intelligence coordinate nine distinct functions — fed by a three-zone data pipeline and optimized by systematic prompt discovery.

The five layers of intelligence

5
Interpretive Intelligence
Cross-tradition synthesis, novel framings, perspective shifts — seeing patterns the user hasn't seen.
4
Context-Aware Retrieval
Multi-signal scoring — developmental stage, emotional state, coaching themes, session history, cultural context, what has worked before.
3
Knowledge Graph
Relationships between works, concepts, traditions, frameworks. This is where the Throughline pattern lives — meta-nodes spanning traditions: Stoic apatheia → Buddhist dukkha → Frankl's logotherapy → post-traumatic growth.
2
Vector RAG
Semantic search over hierarchically chunked content with rich metadata. Every passage is source-traceable to exact chapter, page and paragraph.
1
Foundation Model
Claude, GPT, Gemini — whichever base model serves best for the task. Breadth without depth.

The medallion data pipeline

High-quality, non-contradictory data is the single most effective way to reduce hallucination. The corpus pipeline is a standard enterprise medallion architecture adapted for wisdom content — expensive curation happens once, offline, and is amortized across every conversation.

Raw landing
Bronze Zone
Authoritative copy of every source in its original format. Every book, paper, transcript, ceremony record. No transformation. The audit trail.
Cleaned & structured
Silver Zone
Markdown conversion, hierarchical chunking, contradiction resolution across traditions, 45-column tagging, knowledge-graph edges, throughlines. The expensive work, done once.
Agent-ready
Gold Zone
Per-agent slices. Sasha Core gets Relational Fit metadata. The Knowledge Agent gets the full graph. The Living Portrait gets developmental trajectory data. Optimized for each retrieval pattern.
Processing a book through the full tagging pipeline costs roughly $0.17–$0.30, once. At query time, conversations are fast because the heavy lifting already happened.
Systematic prompt optimization

DSPy, not hand-tuning

Manual prompt engineering is a starting point, not a discipline. Instructions and few-shot examples are discovered programmatically by Stanford's DSPy framework: you provide a seed prompt and a scoring metric, the optimizer tests thousands of variations against the metric and returns what measurably works.

This is an active area of research. Optimizers are better than hand-tuning, not yet a solved problem. We name that honestly.

Reinforcement from real outcomes

The learning loop

Every coaching interaction teaches the system which wisdom nodes produce breakthroughs for which personality structures, in which ontological configurations, under which present-moment conditions. Aggregated, anonymized, governed — this is the data asset no competitor can replicate by buying more models. The corpus gets more precise with every human it serves.

Nine functions, orchestrated by Sasha Core

These are nine functions, not nine branded "agents." The industry has a habit of inflating agent counts — "agent washing" — by naming every function as a separate product. We don't do that. The underlying architecture distributes these functions across a coordinated set of model calls, retrievers, and services; the count is a design choice we'll refine as the system matures. What matters is that each function actually runs, is independently evaluable, and hands off cleanly.

Sasha Core · orchestration
Voice, empathy, continuity of relationship. Coordinates the other functions.
Living Portrait
Who this person IS — constitutional, pattern, present-moment.
Wisdom & knowledge
Retrieval, throughlines, cross-tradition synthesis.
Growth orchestration
Goals, commitments, trajectory over months and years.
Evidence & measurement
Outcomes, evidence packs, measurement cascade.
Relationship intelligence
Family, team, manager, partner — the relational field.
Somatic & wellbeing
Wearables, sleep, stress, nervous-system state.
Coach augmentation
Makes human coaches measurably better — prep, notes, pattern recognition.
Organizational intelligence
Cohort-level patterns with k-anonymity; never individual surveillance.
05 — In daily life

Wisdom shows up where a life is actually lived.

A Wisdom Corpus that only answers philosophical questions is a library. One that meets a person in the Monday-morning moments — the ten-minute 1:1, the first difficult conversation, the hard call before a board meeting, the 3 a.m. identity question — is an intelligence. Below: five different moments, same substrate.

These are Sasha-led interactions — often supported by a human coach on the other side of the relationship. In each moment, Sasha does one of three things: asks a question that reframes, offers a framework small enough to use in the next ten minutes, or simply holds presence. Never tells the person what to do. The short dialogue snippets below are representative — actual wording adapts to voice, culture, and moment.

3 a.m.
A person in an identity question

"I don't know who I am anymore. My marriage is fine. My career is fine. I'm not fine."

Them — I feel like I'm losing my mind.
Sasha — What if this isn't losing your mind — what if something you used to organize your life around is quietly no longer true?
Them — …that's exactly what it feels like.
Sasha — A lot of people hit this in their late thirties. The old self finishing. The new one not here yet. I don't want to fix it for you. I want to stay here with you while it's happening.
Drawn from
  • ·Kegan Order 3→4 transition (developmental stage)
  • ·Meaning Dissolution — transformation mechanism 4 (Frankl, Kuhn-for-the-self)
  • ·Threshold / Initiation dynamics (van Gennep, dark night)

Not problem-solving. Not reassurance. The recognition that some thresholds require a fundamentally new relationship with identity — and the presence to stay in it until morning.

7:00 a.m.
A business analyst, first difficult conversation

A business analyst has to push back on a senior engagement manager's instruction for the first time. Heart rate elevated, sleep was bad, meeting is in 45 minutes.

Them — I have to push back on my manager and I'm terrified.
Sasha — Your nervous system is in alarm. First: three slow exhales, longer than the inhales. Do them now, I'll wait.
Them — …okay. Better.
Sasha — Four sentences. Observation — 'in the draft, the assumption is X.' Feeling — 'I'm worried this lands the client wrong.' Need — 'I want the final to hold up.' Request — 'Can we walk through my concern for five minutes?' That's it. You're not fighting her. You're giving her data.
Drawn from
  • ·NVC 4 Steps (Rosenberg) — a framework they can actually use in the next 45 minutes
  • ·SCARF model (David Rock) — so they understand their own threat response
  • ·Polyvagal grounding (Deb Dana) — wearables show sympathetic activation

Not a lecture. A two-minute prep, a tool small enough to use today, and a brief reflection afterwards that makes the next one easier.

10:30 a.m.
An engagement manager running a 1:1

An engagement manager has ten minutes with someone whose work has dropped off. She doesn't know why. She has done hundreds of these — this one feels different.

Them (to Sasha, pre-meeting) — I have 10 minutes with a teammate. Last week was not like her. I want to help without ambushing her.
Sasha — Notice the pull to diagnose. Try a Learner opening: 'I'm noticing some shift — I might be wrong. How are you really?' Then be quiet for ten full seconds. Most people fill silence with the truth.
Them (in meeting) — I'm noticing a shift. I might be wrong. How are you really?
Teammate — …my mom was hospitalized last Thursday.
Drawn from
  • ·Choice Map (Marilee Adams) — Learner questions vs. Judger questions
  • ·Relational Witness — transformation mechanism 5
  • ·Four-zone privacy — what she can know, what stays sealed

Sasha doesn't give her a script. It reminds her of something she already knows — ask, don't diagnose — at the exact moment she needs the reminder, and then gets out of the way.

2:00 p.m.
A senior partner, before a client decision

A senior partner has to recommend a restructuring that will end careers. He has done this before. This time, a gut-level unease he can't name.

Them — I've made this call a hundred times. Something feels off and I don't know what.
Sasha — Three minutes. What would need to be true about this recommendation for you to feel clean about it?
Them — …that the people losing roles aren't absorbing someone else's bad bet.
Sasha — That's a values-level question, not an analytical one. Your unease is information. You don't have to change the recommendation. You may need to change what you ask the client to see before they accept it.
Drawn from
  • ·Conscious Leadership — Drama Triangle → Empowerment
  • ·Tillich's ultimate concern (Ontological Layer 1)
  • ·Cook-Greuter / strategist-stage framing of values-level conflict

Not moral outsourcing. A clarifying conversation that helps him locate himself before he locates the decision. The decision still belongs to him.

9:00 p.m.
A parent, after a hard evening with a teenager

The second fight this week. Door slammed. She doesn't know if she's doing any of it right.

Them — I yelled. Again. I can't do this.
Sasha — You can, and you're not going to tonight. The research is clear about one thing: it's not the rupture that predicts outcomes. It's the repair.
Them — She won't talk to me.
Sasha — Not a fix. Just a knock on her door, no entry: 'I got it wrong earlier. I'm not asking you to come out. I just want you to know I know.' Then leave. Tomorrow morning is a different day.
Drawn from
  • ·Attachment theory (Levine, Johnson) — relational repair is the health signal
  • ·Gottman's repair attempts — predictive of long-term outcome
  • ·IFS parts work (Schwartz) — which part of her is reacting, which of her child

Not a parenting tip. A different quality of attention. A practice she can do in the next ten minutes that makes tomorrow morning possible.

Coach augmentation

Human coaches working on ReGenesis see the same substrate from the other side. Before a session: the client's Living Portrait, pattern-level hypotheses drawn from the corpus, frameworks likely to land. During: a quiet companion surfacing a phrase, a poem, a question. After: pattern recognition that took ten years of practice to see unaided. The corpus makes human coaches measurably better — not replaced.

06 — Why it matters

Google indexed the world's information.
We're indexing the world's wisdom.

Foundation models are commoditizing. In five years, every company will have access to roughly the same raw intelligence. The moat of a developmental intelligence platform isn't the model — it's what the model is grounded in.

The Wisdom Corpus is governed, curated, structured as a living knowledge graph, and enriched continuously by what actually helps real people grow. That takes years to build. It cannot be scraped. It cannot be replicated by a larger compute budget.

Curated signal
More than 18,000 carefully chosen works, not the entire internet.
Structured relationships
A knowledge graph with throughline meta-nodes — not just embeddings.
Developmental awareness
Tagged for who can receive what, at what stage, through which expression.
Source traceability
Every insight traceable to exact page, paragraph, quote.
Learning loop
Compound improvement from every conversation, across cultures and contexts.
Non-negotiable commitments
—
Optimized for flourishing, not usage time. Sasha nudges toward real human connection and healthy boundaries. Dependency risk is a core safety domain.
—
Radical source traceability. Every meaningful insight is traceable to evidence, or clearly labeled as hypothesis.
—
Four-zone privacy, by architecture. Individual–Sasha vault has no human access. Shared zones are consent-gated. Aggregate zones require k≥10.
—
Not therapy. Not medical care. Coaching, reflective inquiry, behavior-change scaffolding. Escalates when risk arises.
—
Humble about expression. Every wisdom node is a finger pointing at the moon. The system knows the difference.
—
Governed corpus. Cultural and traditional material is curated with attribution, consent where required, and clear limits about where it applies.
Mission

"Put humanity's deepest wisdom in service of every person's growth."

A world where the support to grow — to see clearly, to lead wisely, to live fully — is not a privilege but a given. Where technology amplifies the best of our humanity. Where organizations flourish because their people do.

About ReGenesisSecurity & privacy