A Guided Self-Inquiry App Built on 26 Years of Practice

We all carry a HIPPO: Hidden Internalized Patterns, Perceptions and Old beliefs. The ones that kept you safe as a child are still running, and they are running your adult life.

QuestionsHeal is a guided self-discovery app. It asks questions until you find the belief underneath, then walks you through setting it down. No advice. No chatbot telling you what you want to hear.

QuestionsHeal is not therapy, medical treatment, or a substitute for professional care.

We have made a powerful, consistent, scalable form of guided self-inquiry, accessible around the world.

Dr. Moshe's Story: How the Vis Dialogue Was Born

Tom Meloche, co-founder and CTO, arrived at the same insight from the other direction. He picked up a book claiming asthma could be reduced in 30 days, and tried the one small experiment it suggested: when symptoms start, pause and notice your thoughts before medicating. By day 21 he had stopped reaching for his medication. Later he noticed a second pattern, that letting his thoughts spin into rage brought on back pain. When he changed how he held the situation underneath, the pain eased.

TOM'S STORY: What Happened When I Started Paying Attention

QuestionsHeal launched publicly on April 19, 2026. This is a deployed product, not a prototype.

1. 800+ app users since launch
2. 12 languages live in voice and text: English, French, Spanish, Italian, Hindi, Russian, German, Hebrew, Arabic, Japanese, Mandarin, Portuguese
3. 2,387 automated tests passing in production
4. Voice mode in Dr. Moshe's own voice, with four voice options in every language
5. Cross-session memory, Stripe payments and an admin review dashboard, all live
6. Three safety detector classes running on every session

$19.95/month. 3 Dialogues. 100% money-back satisfaction guarantee.

What people say after one session

"Our last session was really impactful! The back pain I had been experiencing has been reduced to almost nothing. Wow! I can see the great potential this dialogue has for impacting people who want to heal and change but are fumbling around trying to find a way to do that. For me it feels like the missing link." Allie G.

"It got clear down to the core belief. I recognized I was already complete." Elizabeth Jane Taylor

"I was surprised by how quickly it created a sense of depth and connection." Ilana B, Naturopathic Doctor

"It got me to a really unexpected place. All in all, I was really impressed." Lori Kolle, Psychotherapist

"Q was helpful in unraveling a situation I couldn't sort out inside myself. Bringing clarity." Anya

"Dr. Moshe's work is revolutionary, the most powerful healing I've ever experienced or shared. I'm so grateful for the tools I've learned, and cannot recommend a teacher or a healing more effective." Dr. Moore, ND

Individual experiences, shared with permission. Not therapy or medical treatment. Results vary.

The answers are not out there

More advice exists than ever and less of it fits. Endless information, endless experts, no way to tell what is true for you.

The hard part was never access to information. It is being asked the right question, in the right sequence, to reach the answer already inside you. That skill is scarce, expensive, and gated behind a 48-day average wait.

1. 48 days, average wait for a first appointment
2. $150 to $300 per session, where care exists at all
3. 46% of Americans live in a provider shortage area

Limiting beliefs quietly shape how people experience stress, emotion and physical discomfort. Beliefs like "I have to be perfect," "I don't matter," or "I'm not safe" sit mostly outside everyday awareness while influencing behavior, relationships and the body's stress response. Most people never examine them directly. They experience the downstream effects instead: anxiety, burnout, tension, pain, recurring struggle.

Most false beliefs do not feel like beliefs. They feel like a gut certainty, a body alarm, or just reality. They were learned early, reinforced by repetition, and stored as patterns. You can be a capable adult and still be running a rule you installed at five years old.

Nobody had a tool for finding them. At a local entrepreneur gathering a presenter asked how many people in the room knew they had limiting beliefs. Around 50 people were there. Every hand went up. Later that day Tom asked two of them privately whether their beliefs had ever affected their health, and whether things improved when a belief changed. Both said yes to everything, then wanted to tell their stories. The awareness is already there. What nobody had was a reliable way to examine a belief when distress shows up.

Q is the name of the Agent of QuestionsHeal.

When the right questions arrive in the right order, hidden rules become visible, often quickly. Once a false belief is seen clearly it starts to lose its hold, because a false belief cannot survive being seen for what it is. Nothing is added and nothing is fixed. What was always there is uncovered.

Underneath, QuestionsHeal runs a structured multi-agent architecture: a state machine, emotion-sensitive routing, and a retrieval layer built on 1,500+ recorded sessions. It is a questions-based, belief-hunting engine.

A newer, storybook "HIPPO" style presently also exists for the app.

$19.95/month. 3 Dialogues. 100% money-back satisfaction guarantee.

We all know how the generic chatbots echo what we want to hear and offers feel-good validation.

Our Philosophy: Tech That Helps See You're Free, Not Keep You Hooked

Our philosophy: technology that frees you rather than holds you. The business model is not retention at all costs. The goal is that someone sees the belief, sets it down, and gets on with their life. If we become unnecessary to a user, they are free, and that is the point. The door stays open if they need us again.

Three forces are converging. Demand has permanently shifted, with burnout and anxiety at record levels and no sign of returning to baseline. Language models have only just reached the threshold this method requires, where an AI can sustain nuanced, emotionally aware dialogue. And stigma is dissolving fastest among the highest-need groups, with 68% of Gen Z open to AI-assisted personal growth.

Five moats that compound

1. A 26-year method, mapped to 187 numbered dialogue nodes and 284 labelled transitions. It cannot be scraped or bought.

2. A proprietary corpus, calibrated on 184,000+ real interaction patterns from 1,500+ sessions.

3. Multi-agent architecture: 54 trigger classes, 14 pattern interrupts, 4 monitoring agents.

4. AI that asks and never tells. Every insight comes from the user.

5. Owned outright. The prompt system, model routing, safety architecture, evaluation engine and session cost tracking belong to the company, assigned by both founders.

The Vis Dialogue and the 26-year corpus are licensed from the founder under a perpetual, irrevocable, worldwide, sublicensable license. Only the company can end it. Everything built on top, the company owns.

Safety by design.

Every session runs through three real-time detector classes: signals of suicidal ideation, signals of intent to harm others, and disclosures of serious medical risk. When a signal fires, the user is addressed directly, referred to human help, the session stops, and the founders are flagged. QuestionsHeal is a self-discovery tool, not a medical device and not a substitute for emergency or crisis care.

How we make money

Users pay $19.95 per month for 3 Dialogues, with a 100% money-back satisfaction guarantee. Finish one Dialogue end-to-end and request your first $19.95 back in full within 24 hours of finishing, before starting another. Cancel anytime with one click.

1. Subscription: $19.95 a month for 3 Dialogues, with a 100% money-back satisfaction guarantee.
2. Cancellation: Cancel anytime with one click.
3. Refund: Finish one Dialogue end-to-end; request the first $19.95 back in full within 24 hours of finishing, before starting another. No form, phone call, or explanation required.
4. Enterprise: $120 per employee per year with four Dialogues a month included, priced below the $150 to $200 that employee assistance programs cost. Profitable at 50% utilization, where typical program usage is 3% to 8%.

API cost runs about $4 per subscriber per month at expected usage, against the $19.95 they pay. The pay-first model includes a 100% money-back satisfaction guarantee.

Two more channels, both low cost. Gifting lets members buy Dialogues or subscriptions for people they care about, which is distribution at no acquisition cost. Partnering pays practitioners, podcasters and creators per paid user, which turns trusted voices into a sales force.

Where the repeat revenue comes from.

One session can resolve one belief and the patterns attached to it. People carry layers. Someone might resolve a belief about work pressure now and come back months later for one about a relationship. The win earns the next session. The model runs on trust and results.

Our strongest channel so far is affiliate partnering.

Influencers, practitioners, podcasters and creators promote QuestionsHeal to audiences that already trust them, and we pay them per paid user. It costs nothing up front, it scales with the size of their following rather than with our headcount, and of everything we have tried it is producing the best results to date.

How we are raising

We are raising $1.9 million in four tranches, not as one lump sum. A tranche is a fixed amount of money with a fixed goal. We raise it, spend it, hit the goal, and only then raise the next one. Each tranche is its own full budget. Nothing carries over. Each gate is a fresh decision made against real numbers, not a promise made a year earlier.

This round is tranche 1: $150,000 on a post-money SAFE, $10M valuation cap, with a 20 percent discount. Tranches 2 to 4 convert at a $12M cap with the same discount.

1. Tranche 1: $150,000, months 1 to 4. Goal: make one subscriber profitable.
2. Tranche 2: $250,000, months 5 to 9. Goal: prove trust beats paid ads, and cross the self-sustaining line.
3. Tranche 3: $500,000, months 10 to 15. Goal: back the two channels that won, and open enterprise.
4. Tranche 4: $1,000,000, months 16 to 21. Goal: scale what is proven, and hold a reserve.

Three rules hold across all four. Trust first: nobody pays for something they have not seen and do not yet believe in. One profitable subscriber before many: growth against unknown churn burns money into a hole you cannot see. Prove the English-speaking market before any new language.

Tranche 1: $150,000, months 1 to 4

The goal is to make the business readable. Reach 250 paying subscribers still active at month 4, across two or three channels, so churn and cost per customer stop being guesses, and start building our own audience.

Where the $150,000 goes:

1. 30% Meta ads into the trust funnel. A measured cost per paying subscriber. $45,000.
2. 8% Marketing and content direction, $3,000 a month: runs the funnel, the clips and the weekly tests. $12,000.
3. 10% Video, content and a marketing assistant. One new video, one landing-page change and one test every week. Posting and list building. $15,000.
4. 3% More tests in the trust funnel. $4,000.
5. 8% Dialogue engine development, $3,000 a month: builds and runs Q's dialogue engine. $12,000. 7% Tech. One rig for Tom and more memory for the testing machine that runs the clip and marketing pipeline. $10,000.
6. 3% A part-time engineer who learns the code. The first step on key-person risk. $5,000.
7. 3% Outside code review. A live engineer reads the code every month as AI writes more of it. $4,000. 3% Administration and investor reporting, $1,000 a month: books, filings and investor reports. $4,000.
8. 5% Software, tools and customer interviews. AI subscriptions, cohort reports, cancellation reasons, cost per subscriber by channel, and 25 to 30 structured interviews. $8,000.
9. 5% Diligence and books. Insurance quotes, an outside security review, bookkeeping and a monthly close. $8,000.
10. 6% Existing bills. Cooley LLP, our incorporation counsel. $9,305.
11. 9% Reserve. Carries the company past month 4 while the next tranche is decided. $13,695. The decision at month 4 is written down in advance. 250 or more paying subscribers still active: the gate is met and tranche 2 can open. 200 to 249: the board decides. Under 200: tranche 2 does not open, and we fix the funnel first.

The decision at month 4 is written down in advance. Under 188 subscribers: stop and fix the funnel before raising more. 188 to 262 subscribers: founders and investors review the numbers together. 263 to 375 subscribers at $120 to $175 each: raise the next tranche. More than 375 at under $120 each: raise the next tranche sooner.

Where the $250,000 goes:

1. 16% Meta ads, held as the benchmark every other channel is measured against. $40,000.
2. 16% Our own audience: a part-time social media person, paid promotion of our own clips, editing and production. $40,000.
3. 6% Marketing and content direction, $3,000 a month for five months. $15,000.
4. 12% A part-time growth partner who recruits practitioners and lines up podcasters and influencers. $30,000.
5. 8% Influencer and podcast activation, paid as a share of the revenue they bring in. $20,000.
6. 4% Marketing assistant, made steadier. $10,000.
7. 6% Dialogue engine development, $3,000 a month for five months. $15,000.
8. 10% Retention product work: the features that make people stay, measured against the churn number tranche 1 produced. $25,000.
9. 6% Contract engineering, a second pair of hands on the code. $15,000.
10. 4% The question library to 150 or more pages, and the insider circle launch. $10,000.
11. 10% Code reviewer, legal, accounting and tools. $25,000.
12. 2% Administration and investor reporting, $1,000 a month for five months. $5,000.

The gate to tranche 3: $20,000 in monthly recurring revenue. 10 percent net monthly growth. 100,000 followers. 40 percent of new subscribers arriving through channels we did not pay for. 50 or more practitioners sending real referrals. A first clean read on six-month retention. And a stated answer to the question the whole strategy turns on: content and referral versus paid ads, which one wins.

Tranches 3 and 4: direction, rewritten at each gate

Tranche 3, $500,000, months 10 to 15. Grow from 1,000 toward 1,500 subscribers. Put most of the acquisition money behind the single cheapest channel for a retained subscriber. Hire the first enterprise salesperson. Build retention features and a second dialogue technique. Bring a second engineer toward half time. Add insurance. The gate: 1,500 subscribers, two channels at a predictable cost per retained subscriber, six months of clean retention data, and one paid enterprise pilot that renews.

Tranche 4, $1,000,000, months 16 to 21. Scale what is proven and resolve what is not. $400,000 is held back on purpose and spent on whichever channel has the lowest cost per retained subscriber. The rest goes to direct marketing at real weight, the team, closing the first enterprise contract or stopping, the practitioner network past 150, and one second-market pilot.

The end state we are aiming at: two channels we would defend to any investor. One we would kill. Clean data on the rest. A following an acquirer can count, and a body of dialogues an acquirer cannot build.

Our approach is AI-first and human-supervised. AI does the heavy lifting across product, content and marketing, and we pay people to review and refine. Headcount grows only as revenue and users justify it.

NVIDIA Inception is a prestigious program designed to help startups evolve faster through access to NVIDIA's technology, expertise, and go-to-market support. Acceptance signals that NVIDIA recognizes QuestionsHeal™ as a serious AI-driven company building at the cutting edge.

© 2025 NVIDIA, the NVIDIA logo, and NVIDIA Inception are trademarks and/or registered trademarks of NVIDIA Corporation in the U.S. and other countries.

$19.95/month. 3 Dialogues. 100% money-back satisfaction guarantee.

Global Reach: A Twelve-Language Footprint

Twelve languages, one engine.

English, French, Spanish, Italian, Hindi, Russian, German, Hebrew, Arabic, Japanese, Mandarin, Portuguese

This expands the company's addressable market in a structurally important way. A user-facing platform that only operates in English can serve only a fraction of the people it could otherwise help — and competes inside the most saturated personal growth market in the world.

Twelve languages cover roughly two-thirds of the world's speakers by first or second language, including every G7 economy and several of the largest underserved growth markets in personal development tooling. Critically, most of these markets have no English-language wellness incumbents with localized footholds.

Operationally, the cost of adding each additional language is bounded — voice models, prompt review, and right-to-left interface work for Hebrew and Arabic have already been absorbed. The marginal cost of serving a Spanish-speaking user is functionally identical to the marginal cost of serving an English-speaking one. Twelve languages, therefore, do not multiply our operating cost; they multiply our reachable market.

Speak one of these languages? $19.95/month. 3 Dialogues. 100% money-back satisfaction guarantee.

Frequently Asked Questions

How is my investment structured? What does a SAFE with a $10M cap mean?

A SAFE gives you the right to shares in the future. The cap sets the highest price your money converts at. This tranche, the first $150,000, converts as though the company were valued at $10M, however high the next round prices. Tranches 2 to 4 convert at a $12M cap. Every SAFE in this raise also carries a 20 percent discount on the next priced round. If a future round prices lower than the cap, you get the better of the two.

If a single session resolves something, where does repeat revenue come from?

People carry layers of beliefs. One session can settle one belief and the patterns attached to it. Someone who resolves a belief about work pressure often returns months later for one about a relationship, because the first result earned the second visit. The subscription is sticky because it works when it is needed.

How do you handle privacy?

Sessions are personal, so privacy is built into the design. User data is anonymized and encrypted, and we will never sell it. Revenue comes from subscriptions, not advertising. We avoid invasive technology such as facial emotion recognition.

What makes "26 years of experience" more than a story?

The 26 years produced a system, not just a track record. The method is mapped to 187 numbered dialogue nodes and 284 labelled transitions, calibrated against 105,800+ interaction patterns drawn from more than 1,500 recorded sessions. Facilitators trained in the method see the same results in their own practices, which is what makes it reproducible and teachable rather than personal talent.

How is this different from a general AI chatbot?

General chatbots are built to be agreeable and to supply information. They tend to tell you what you want to hear. Q is built to ask, following a sequence developed over 26 years, and it will ask questions you would rather avoid. One gives comfort. The other gives insight.

How do you know it works?

Dr. Moshe's 26 years of practice produced documented belief shifts and reported improvements in wellbeing, reproduced by the facilitators he trained. Early app users report meaningful shifts inside a single session, including feeling more at ease physically. Results vary, and we are publishing early metrics on retention, session length and reported outcomes as the cohort grows.

Is the app live? Can I try it before I invest?

Yes to both. QuestionsHeal launched publicly on April 19, 2026. Go to q.questionsheal.com and start with the $19.95/month plan: 3 Dialogues, with a 100% money-back satisfaction guarantee.

Is this a medical device?

No. QuestionsHeal is a self-discovery and personal development tool. We do not diagnose or treat conditions, and it is not a substitute for professional care or crisis support.

What happens in an acquisition?

If the company is acquired before your SAFE converts, the SAFE pays out according to the deal structure. If conversion has already happened, you hold shares like any other holder.

What is the timeline to liquidity?

Venture-backed companies typically take seven to ten years. Our path runs through a Series A within about 18 months, with acquisition interest possible earlier.

Why is the timing right?

Awareness of emotional wellbeing is high and trust in existing options is low, which is an unusual combination. Language models only recently became capable of sustaining this kind of dialogue. The window to define this category is open now, and generic AI wellness apps are filling the space quickly.

What is the long-term vision?

A global self-discovery platform serving individuals, employers and public programs in their own languages. Twelve languages are live today. Capturing 1% of a $537B market would build a company at meaningful scale, and the mission is worth more than the multiple.

Q: Is this just another tech company using my data? How do you handle privacy, especially with something so personal?

A: We are a human-first company, a tech company second. Privacy is at the core of our design. All user data is anonymized and encrypted. We will never sell user data. Our business model is based on subscriptions, not advertising. We are also avoiding invasive technologies like facial emotion recognition to ensure users feel safe and in control. Your trust is our most important asset.