Product Breakdown

What powers the analysis
your competitors can't access

Not a ChatGPT wrapper. A collision engine backed by 25,000+ practitioner claims, three cognitive modes, and a self-improving knowledge base.

25K+
Knowledge Claims
3
Cognitive Modes
13
Source Channels
<5 min
To Artifact

01 / The Problem

Generic AI gives generic answers

The difference between a ChatGPT response and a Pharallax analysis.

Generic AI

  • Trained on the whole internet. Knows everything, understands nothing
  • Gives "focus on your customer" advice
  • No structured output. Just paragraphs
  • Same answer every time regardless of context
  • Can't tell you what you're NOT seeing

Pharallax

  • 25K curated claims from practitioners who've done the thing
  • Names the specific mechanism, not the category
  • Structured artifact: blind spots, leverage moves, risks
  • Three cognitive modes surface different insight classes
  • Built to find what you're not discussing

02 / The Knowledge Engine

25,000+ structured claims from practitioners

Not scraped. Extracted, structured, and indexed from 1,000+ hours of practitioner content.

Extraction

Every claim is extracted from video transcripts using AI - not summarized, extracted. "Hormozi says pricing should anchor to pain" becomes a structured, searchable knowledge claim with source attribution.

Indexing

Claims are embedded as vectors in a semantic search engine. When you submit a problem, the engine finds the 5 most relevant claims across all 25K - regardless of which channel or speaker they came from.

Collision

The Collision Engine forces synthesis between distant fragments. A pricing insight collides with a distribution pattern collides with a behavioral economics principle. Cross-domain, not linear.

03 / Cognitive Modes

Three lenses. Different insight classes.

Each mode uses a different reasoning framework from published cognitive science.

Premortem

"Assume it failed. Work backward."

One year from now, your strategy failed. This mode performs the autopsy. What killed it? What did you refuse to see? Surfaces structural weaknesses before they become expensive.

Dialectic

"Two minds argue your problem."

Mind One sees only opportunity. Mind Two sees only risk. They can't merge. Where they agree, that signal is 10x more valuable. The tension between them IS the insight.

Reflexive

"Your assumptions are the problem."

Maps feedback loops between what you assume and what you see. Finds the one belief that, if changed, restructures everything downstream. The leverage belief.

04 / Process

Problem to artifact in under 5 minutes

Four phases, streamed in real-time.

01

Submit

Describe your strategic problem in plain language. Pricing, positioning, market entry - anything with stakes.

02

Collision

Your problem hits 25K+ claims. Semantic search finds relevant fragments. Collision engine forces cross-domain synthesis.

03

Analysis

A cognitive mode runs against the collision output. Streamed token-by-token so you see the thinking unfold.

04

Artifact

Structured report: blind spots, highest-leverage move, unspoken risks, concrete next action. Downloadable as PDF or Markdown.

05 / Sample Output

What the artifact looks like

Real output from a dialectic mode analysis on a SaaS positioning problem.

PARALLAX ANALYSIS COMPLETE
Mode: Dialectic - opposing forces | 5 knowledge claims
Blind spots surfaced
  • Treating referral behavior as identical across segments when multi-room studios likely aren't championing the product - they're renting it until replacement arrives
  • Assuming feature requests equal willingness-to-pay when studios may want you to become their free R&D department
  • Framing this as "upmarket vs downmarket" when it's actually "word-of-mouth business vs. sales-motion business with unknown CAC"
Highest-leverage move

Segment revenue by referral source this week to identify which customer type actually drives acquisition - if solo producers refer and studios don't, the upmarket opportunity is a retention trap, not a growth engine.

Risk not being discussed

With 11 months runway and 3 months of flat growth, you're approximately 4-5 months from a forced decision where panic drives strategy instead of data.

Next concrete action

Send a fake door email to all multi-room studios today offering premium features at a specific price point. Count binding commitments within 72 hours.

06 / Auto Research

The engine improves with every session

Every analysis you rate trains the knowledge engine to deliver sharper results.

Ratings
Every session ends with a 1-5 rating. Scores below 5 prompt written feedback on what was missed or wrong.
Gap Detection
Low-rated sessions flag knowledge gaps. If the engine can't find relevant claims for a topic, that gap is logged for sourcing.
Claim Scoring
Knowledge claims that appear in highly-rated sessions get reinforced. Claims in low-rated sessions get deprioritized. Signal rises, noise falls.
Expansion
Detected gaps trigger sourcing of new practitioner content. The knowledge base grows in the directions users actually need, not randomly.

Your competitors are making decisions with less.

First analysis is free. No credit card.

Run Your Free Analysis