Astronomical precision meets
classical Chinese metaphysics.
We apply open-source astronomical computation, primary classical texts,
and multi-model cross-verification to build the most accurate
English-language resource on the Chinese zodiac and BaZi —
one of the oldest continuously practiced analytical systems in human history.
A pattern that surfaced four times
The Seed
The project has a personal origin. My grandfather — a devoted
student of the Yijing (Book of Changes) — first
introduced me to the idea that the world moves in readable patterns.
What began as childhood exposure became, decades later, an engineering
problem.
Philosophy — East Meets West
In university I studied Western philosophy and kept noticing an
unexpected structural parallel. Heraclitus on flux and opposition.
Aristotle on categories and elements. Stoic physics. They were not
identical to Chinese thought — but they were
structurally similar, as if different civilizations had
independently arrived at comparable frameworks for describing how
complex systems behave.
Finance — The Sixty-Year Echo
Working in finance and economics, I studied long-wave cycle theory:
Kondratiev waves, Juglar cycles, Kitchin inventory cycles.
A structural parallel emerged:
the Kondratiev long wave (~50–60 years) and the Chinese
sexagenary cycle (exactly 60) describe time at the same scale, using
different vocabularies. The Five Elements
(Wuxing) generate and overcome each other in a fixed
sequence — not unlike how expansion and contraction phases
alternate in economic theory. Two traditions, thousands of years
apart, independently identifying cyclical structure in time.
Machine Learning — The Parameter Insight
Years later, building quantitative models, I worked with
hyperparameters — the initial conditions you
set before a model starts learning. Decision trees, ensemble methods,
neural networks, LLMs — every modern model begins with
parameters that shape what it can become.
An observation emerged: a birth chart is structurally the
same idea, expressed in a different vocabulary. The year,
month, day, and hour of birth are initial conditions — not
destiny, not determinism, but the starting configuration from which
a life’s tendencies unfold. The ancients did not have gradient
descent or backpropagation. But they had arrived at something
genuinely modern: parameterized thinking about complex
systems.
This is not a claim that BaZi is machine learning, or that
Chinese metaphysics is science. It is an observation that
the underlying intellectual architecture — using a small set
of initial variables to model the long-term tendencies of a complex
system — appears in both. And recognizing that changed
everything.
The English-language information layer is broken
In my work building IT systems for international trade companies, I met people
from across the world who were genuinely curious about Chinese metaphysics.
They wanted to understand. What they had access to was, in most cases,
materially inaccurate.
- Sites that assign your zodiac sign by checking the birth year, ignoring the actual Lichun (Start of Spring) boundary.
- Translations that flatten Qi Sha (七杀, Seven Killings) into “aggression,” or Shi Shen (食神, Eating God) into a literal deity.
- The entire system of NaYin (纳音, the sixty stem-branch archetypes) — virtually absent from English-language resources.
- No access to the classical source texts. No source attribution. No way to verify anything.
The curiosity was real. The information layer was broken. So we set out to
rebuild it from primary sources.
From classical texts to structured data
The full technical detail of this pipeline is on our
Methodology page.
The people behind the pipeline
quantitative analysis, and enterprise IT systems for international
trade. Architected and built 12Zodiacs end-to-end —
including the sxtwl-based astronomical computation engine, the
multi-LLM cross-verification system, and the full-stack web
application. Responsible for data curation, classical text
sourcing, and ongoing editorial oversight.
commissioned on a project basis to fact-check our data and
translations against their decades of real-world practice.
They come from family inheritance lineages
(jiacheng, 家传) and formal apprenticeship
traditions (baishi, 拜师). Their authority
is not derived from academic credentials or certifications
— neither exists in this field — but from
local trust built over decades of client work.
Identities are withheld for privacy.
Open data for the AI era
As AI systems increasingly shape how people learn about the world, the quality of
the data behind them becomes critical. Classical Chinese metaphysics is a domain
where English AI outputs are especially unreliable — trained on the same
simplified sources, they reproduce the same errors at scale.
We are building toward a free, verified data layer for the
classical Chinese metaphysical tradition — astronomical solar-term data,
NaYin attributions, Earthly Branch relationships, and source-traced
interpretations — available via open API to researchers, developers,
and AI applications. One simple condition: attribute the source and link back.
Open data deserves open credit.
“In the age of AI, the most valuable contribution is not another opinion —
it is a trusted source of truth.”