Zypher
Model your next move
8 followers
Model your next move
8 followers
Zypher is an AI decision engine inspired by ancient royal modeling systems. We don’t do vague advice. Every dilemma is processed into three weighted strategic paths, exposing confidence distributions and risk signals. Speculation ends where calculation begins.
Free Options
Launch Team / Built With
Render Sandboxes: a computer for every agentIsolated compute, persistent state and a real filesystem
Promoted
Hey everyone 👋 Steve (@stevekwok) and Fiona (@fiona_tang1) here.
Like most of you here, we build things. We live in code, data, and dashboards.
But our hardest decisions have never been solvable by any of that. When to quit. When to leap. When to wait.
A lot of people turn to 🃏 tarot or 🪐 astrology when they’re stuck. Not as magic. As a thinking tool. Not because they want magic. Because they want a lens. A structure. A way to think.
While building Zypher, we found something we didn’t expect in old Arabic manuscripts: Jafr (علم الجفر). A formal, repeatable system that treats letters as numbers and runs them through layered computations.
We went back to the sources and rebuilt it. Then we paired it with a new layer of AI, not to invent answers, but to translate the computed outputs into language you can actually use.
That’s the idea behind Zypher: a collision between classical algorithms and modern AI. Ancient structure, modern interpretation.
Some people will call this pseudo-science. We get it.
But when you’ve exhausted every argument and still can’t decide, a little structure can be its own kind of certainty.
That’s what we’re building.
---
What is Jafr? And what is Zypher?
🧮 Before algorithms traded stocks, they read fates.
Between the 7th and 13th centuries, scholars in the 🌙 Arabic world developed Jafr (علم الجفر). The science of alphabetical divination.
A structured, repeatable method. You encode a question into the numerical values of the 28 Arabic letters, then transform them through layered math to surface hidden patterns.
👑 Royal courts used it for wars, marriages, successions. Ibn Khaldun dedicated an entire chapter of his Muqaddimah to it. The system traces back to Imam Ali and was formalized by Imam Ja'far al-Sadiq in the 8th century.
We went deep into the classical texts. What we found wasn't mysticism. It was math. Matrices, elemental rotations, positional shifts, interleaving. Real computational pipelines, designed centuries before computers existed.
---
Fast forward to today.
Millions of people ask GPT-4o about their biggest life decisions. Career moves, relationships, timing.
It's good at many things. But ask it "should I take this job?" and you get the internet's most statistically average answer. It hedges. It says "consider both sides." No real framework for your situation.
---
Zypher works differently.
The input is simple: your name, 👩 your mother's name, and your question. That's it.
In the original Jafr tradition, your name and maternal lineage are the seed of the entire computation. They believed your identity already encodes the trajectory of your fate.
Zypher transliterates these inputs into Arabic, converts each letter to its Abjad numerical value, then runs them through a chain of matrix operations. Mirroring, elemental rotation, positional shifts, interleaving.
The output is deterministic. Same name, same mother, same question, same result.
AI comes in at the end. It interprets the letter sequences into weighted outcomes you can actually understand. Three possible paths. Confidence percentages.
This isn't AI making up an answer. It's AI reading one that was already computed.
And you don't just get a verdict and walk away.
Zypher is conversational. 🧵 Ask follow-ups. Challenge the interpretation. Explore what each path means for you. The computation stays fixed.
✨ The conversation is as open as any chat with ChatGPT.
---
🤔 How is this different from tarot, astrology, or Bazi?
Those systems are either random (tarot gives a different spread every time) or fixed at birth (astrology and Bazi tell you who you are, but can't answer "should I take this job on Thursday?").
Jafr is question-native. Your specific question is mathematically part of the computation. Not a lens for reading a pre-existing chart. No shuffling. The result is derived from your identity and your exact words.
---
26 algorithms. Career pivots, complicated relationships, or what your cat's past life looked like. One of them was designed for exactly your kind of question.
We have a free tier. 30 readings per month. No credit card.
Give it a spin. You'll be surprised how specific it gets.
And if you have any feedback, ideas, or just want to say hi, reach out anytime. We'd love to hear from you.
Discord: https://discord.gg/BHYMAn6w
Email: team@zypherit.com
X: x.com/zypherit
Hi Product Hunt, I’m Fiona.
AI is accelerating everything: generation, prediction, optimization. But more output doesn’t always equal clarity.
Over the past two months of testing Zypher, I stopped looking for the “perfect” answer. Instead, I found a level of clarity I hadn’t experienced before, regardless of which path I ultimately chose.
Following trends is frictionless. Breaking old decision patterns isn’t.
What changed wasn’t the chaos around me. It was my relationship to uncertainty.
Fear thrives in the unknown. Zypher makes it visible.
Zypher doesn’t replace judgment. It processes complex dilemmas into three weighted strategic paths, exposing confidence distributions and risk signals.
Zypher sees the structure. You make the move.
I’d love to hear how you handle decision noise in your own work.