// VibeAlchemy Marketing · Landing page function LandingPage() { return (
{/* ─── HERO ───────────────────────────────────────────────────── */}
{window.PRIVATE_BETA !== false ? 'Private beta · invitation only · founding pricing held for invitees' : 'Founding offer · First 100 paying customers — 50% off year 1'}

Validation that gets you to the idea worth building.

12 agents research it, argue with it, and pivot it when it cracks. You leave with a verdict you can defend, every claim sourced. Even if the idea isn't the one you walked in with.

{window.PRIVATE_BETA !== false ? ( { e.preventDefault(); openWaitlist({ source: 'marketing' }); }}>Join the private beta → ) : ( Start free → )}
{window.PRIVATE_BETA !== false ? ( <> Invite-only beta Reply in a few days ) : ( <> Free Spark tier No card )}
{/* ─── TRUST STRIP ────────────────────────────────────────────── */} {/* ─── THE PROBLEM ────────────────────────────────────────────── */}

Done right, it takes a month: market sizing, customer research, competitive landscape, risk analysis, stress-testing the business model. Done wrong, it misses the assumption that actually kills the idea.

So most founders skip it, and build something the market was always going to reject. The ones who don't lose a month they could have spent shipping.

~43%
of VC-backed startups fail due to poor product-market fit. CB Insights · 2024 study (431 VC-backed failures since 2023) →
{/* ─── PIPELINE ───────────────────────────────────────────────── */}
{[ { n: '01', name: 'Frame', icon: StageIcons.Frame, desc: 'Lock the opportunity: who, what, and why now.' }, { n: '02', name: 'Research', icon: StageIcons.Research, desc: 'Find the customers and what they actually want.' }, { n: '03', name: 'Analyse', icon: StageIcons.Analyse, desc: 'Size the market, score the moat, work out the unit economics.' }, { n: '04', name: 'Challenge', icon: StageIcons.Challenge, desc: 'Three skeptics try to break the idea before you do.' }, { n: '05', name: 'Decide', icon: StageIcons.Decide, desc: 'A clear verdict, with the reasoning you can defend.' }, ].map((s) => (
{s.n}
{s.icon}
{s.name}
{s.desc}
))}
go('/how-it-works', e)}>The full pipeline, stage by stage →
{/* ─── WHAT YOU GET ───────────────────────────────────────────── */}
{[ { title: 'Decision Memo', body: 'The verdict, the score, the riskiest assumption, and what to do next.', kind: 'memo' }, { title: 'Market Intelligence', body: 'Bottoms-up market sizing, who\'s already in it, what would defend you, what would pay.', kind: 'market' }, { title: 'Red Team Report', body: 'Three skeptics, three angles, three failure modes you need to answer.', kind: 'red' }, { title: 'Pivot Map', body: "When the verdict isn't a clean Pursue, the stronger angles adjacent to your idea.", kind: 'pivot' }, ].map((d, i) => (
go('/examples', e)}>
{`Doc ${String(i + 1).padStart(2, '0')}`}

{d.title}

{d.body}

go('/examples', e)}>See sample →
))}
{/* ─── BUILT FOR ─────────────────────────────────────────────── */}
{[ { label: 'Solo founder', title: "Exploring an idea you haven't shown anyone yet.", body: "You don't have a co-founder to argue with. The Red Team is the closest thing.", }, { label: 'First-time founder', title: 'Serious about your first venture.', body: "You haven't built the muscle for what kills startups. The memo names it for you.", }, { label: 'Pivoting founder', title: "Figuring out the next angle after the current one didn't work.", body: "You know what didn't work. The Pivot Map shows the adjacent angles worth running.", }, ].map((a, i) => (
go('/for-founders', e)} style={{ cursor: 'pointer' }}>
{a.label}
{a.title}

{a.body}

go('/for-founders', e)}>See more →
))}
{/* All-In as graduated next step */}
go('/for-all-in', e)} > NEXT STEP · ALL-IN
Going all-in on one idea?

Stack All-In on top — investor framing, verifier pass, source citations, share blind with VCs.

go('/for-all-in', e)}>See All-In →
{/* ─── COMPARE ────────────────────────────────────────────────── */}
{[ { name: 'VibeAlchemy', sub: 'Multi-perspective deliberation', us: true }, { name: 'Single-pass validators', sub: 'A score in 60 seconds' }, { name: 'Prompt + chat', sub: 'ChatGPT, Claude in a chat window' }, { name: 'Skills + agents', sub: 'A SKILL.md doc + a general agent' }, ].map((c, i) => { const rows = [ [ 'Pivots, re-frames, re-runs the case on the sharper version', 'Rewrite the input. Pay for the report again.', 'Start a new chat. Lose the thread.', 'No memory of the prior idea. Starts from zero.', ], [ 'A stack of documents you can share and defend', 'A dashboard score and a few bullets', 'A chat thread that drifts off topic', 'Whatever the agent decided to focus on — no standard output', ], [ 'Three skeptics try to break it before you do', 'Optimizes for the upgrade button — your idea scores 74/100', 'Agrees with you. LLMs are trained to please', 'One agent has to play both skeptic and proponent — dissonance, not deliberation', ], [ 'Same structured output every time', 'Varies — same idea can score 60 one day and 80 another', 'Prompt-dependent. Same idea, different answer next time', 'Agent picks angles, depth, and when to stop — different every run', ], [ 'Every claim ties back to a source — share the memo, defend the call', 'Hard to defend from a screenshot of a 74', 'Cannot share a chat as a decision', 'No enforced citations. No audit trail.', ], ]; const kickers = ['When the idea changes mid-validation', 'What you get', 'How it handles a bad idea', 'Consistency', 'Defensibility']; return (

{c.name}

{c.sub}
{rows.map((r, j) => (
{kickers[j]}
{r[i]}
))}
); })}
Full breakdown vs{' '} go('/compare/preuve', e)} style={{ cursor: 'pointer' }}>Preuve {' · '} go('/compare/ideaproof', e)} style={{ cursor: 'pointer' }}>IdeaProof {' · '} go('/compare/validatorai', e)} style={{ cursor: 'pointer' }}>ValidatorAI {' →'}
{/* ─── BUILT BY ───────────────────────────────────────────────── */}
M

Built PyTorch Tabular. Wrote a Packt-published book that ML teams at Amazon and Microsoft use. Now applying that same rigor to the part of building startups that usually gets done badly.

"Most validation tools agree with you. That's confirmation bias with extra steps." Manu · founder
go('/about', e)} style={{ fontFamily: 'var(--mono)', fontSize: 11, letterSpacing: '0.14em', textTransform: 'uppercase' }}>Full story →
1.6k
GitHub stars · PyTorch Tabular
6 + 2
Papers · Patents (peer-reviewed)
2nd
Book edition · Packt
{/* ─── PRICING TEASER ─────────────────────────────────────────── */}
go('/pricing', e)} style={{ fontFamily: 'var(--mono)', fontSize: 11, letterSpacing: '0.14em', textTransform: 'uppercase' }}>Compare full features across tiers →
{/* ─── FAQ TEASER ─────────────────────────────────────────────── */}
go('/faq', e)} style={{ fontFamily: 'var(--mono)', fontSize: 11, letterSpacing: '0.14em', textTransform: 'uppercase' }}>See all questions →
{/* ─── FINAL CTA ──────────────────────────────────────────────── */}
); } const LANDING_FAQ = [ { q: 'Is it live now?', a: "We're in private beta. The full pipeline (Spark, Validate, Founder, All-In) is built and running — we're inviting founders in small batches so we can support each one properly. Request access to get on the list.", }, { q: 'What does requesting access commit me to?', a: "Nothing. Drop your email and where you are with your idea — we invite people as capacity opens up. No card, no payment. When you're invited, you'll get founding-member pricing on any paid tier.", }, { q: 'Will it actually kill bad ideas?', a: "Yes — by design. We're built to find what would break the idea before you do. But a Rethink verdict isn't the end. Every Pivot or Rethink comes with a Pivot Map — the stronger angles adjacent to your original, ranked. The Rethink is the first step, not the last.", }, { q: 'How is this different from ChatGPT, Preuve, IdeaProof?', a: 'Three things: multi-perspective debate instead of one model agreeing with you, three adversarial skeptics whose job is to break the idea, and a structured Decision Memo you can defend to a co-founder or investor. See the full comparison above.', }, { q: 'Will the founding-member price hold after launch?', a: "Yes — for the first 100 paying customers. The 50% off applies to year 1 on any paid tier (Validate, Founder, All-In), and we'll honor the same rate on renewal. Spark is free either way; upgrade within the founding-100 window to lock the discount.", }, ]; window.LandingPage = LandingPage;