The first protocol trading / research / project suite built to serve
No signup required · Try the live demo →Synapse Terminal delivers a fully systematic, deterministic, and mechanical trading intelligence framework engineered to replace discretionary true/false decision-making with mathematically rigorous, data-driven processes.
This framework is purpose-built to preserve and compound trader capital through:
The result is a zero human-in-the-loop, mathematically traceable edge — transforming subjective trading into objective, repeatable, capital-preserving systematic execution.
19 years since 2007 · 55+ GB of clean Nasdaq-100 (NQ) tape from CME Globex — tick, 1-minute and hourly — partitioned by an industry-first time-and-price classification across daily, weekly, and session profiles. zstd-parquet, deterministically gated.
Strict statistical validation backtested across macro regimes. Probability-weighted, p-value tested, zero human-in-the-loop.
Fewer than 30 indexed occurrences in NQ/ES? The indicator is flagged statistically unsafe — preventing false positives by design.
Two Mac Minis run a local Python setup-detection engine 24/7. Bespoke algorithms, razor-sharp mathematical indicators, executed on private M-series hardware — faster, isolated, and fully under your control.
Zero predictive guesswork. 100% mathematical trace — from raw CME transaction tapes, validated on private Apple Silicon, straight to your terminal.
"""
Synapse Edge — ingestion + cleaning pipeline.
Reads raw CSV/parquet exports from Databento.
"""
from pathlib import Path
import pandas as pd
SESSION_BOUNDS_NY = {
"asia": ("18:00", "02:00"),
"london": ("02:00", "07:00"),
"nyam": ("09:30", "11:30"),
"macro": ("09:50", "10:10"),
}
"""
Synapse Edge — time-of-day expectation profile.
For each (session x weekday) bucket compute:
median return, 80th-pct range, reversion hit-rate.
"""
def time_of_day_profile(df):
df["weekday"] = df["ts_ny"].dt.day_name()
return df.groupby(
["session_date","session","weekday"]
).agg(open=("open","first"),
high=("high","max"),
low=("low","min"),
close=("close","last"))
"""
Synapse Edge — Model Weights Loader
Executing on private Apple Silicon (M-series).
"""
import torch, json
from transformers import AutoModel
def load_weights():
print("Initiating tensor distribution...")
for i in range(1285):
pass # MPS load
The same engine — two Apple M4 Silicon local machines, 19 years of clean tape since 2007, and the industry-first time-and-price application — drives every Synapse product. Each product surfaces it differently. Here's how the three core flows work.
You click to claim your hardware-bound license inside your Synapse account. The Pine Script v6 source loads natively in your TradingView layout — session typology markers, weekly bias overlays, kill-zone shading, FVG / OB / SMT divergence triggers — every line wired to the same N ≥ 30 statistical floor that runs in the back-engine. Updates push silently. Razor-sharp mathematical indicators. No black boxes, no discretionary overrides.
Two Apple M4 Silicon local machines run a 24/7 autonomous Python signal-generation agent against a local data engine of 19 years (since 2007) of clean NQ / ES Globex tape. Every candidate setup is gated by N ≥ 30 + t-statistic significance and pushed through bespoke detection algorithms. Triggers stream out sub-50ms to your terminal. Every closed signal is auto-graded by the engine — entry quality, MFE captured, MAE tolerated, replay-vs-live deviation — and the grade lands in your dashboard. Your edge is measured, not estimated.
You log a trade idea, observation, or psych note in Compass — Supabase-backed, timestamped, immutable. The system indexes the idea against the ICT knowledge graph (two decades of Michael J. Huddleston's Twitter Spaces, live streams, and lectures — segmented and topic-tagged). At the right time — NY AM open, kill-zone entry, weekly regime reset, macro release — Compass surfaces a reminder paired with the exact ICT teaching that maps to your logged idea. Sundays, the engine pushes the weekend mirror — your week's logged ideas vs. the recorded outcomes — so the loop closes from idea → execution → review.
Whether you're an individual trader or a business — the same Apple Silicon engine powers both sides. Today, Synapse is laser-focused on Index Futures (NQ / ES / YM); other asset classes are on the roadmap.
Premium custom technical indicators mapping mathematical ranges and session skews natively on your TradingView layout. Fully compatible across NQ / ES / YM front-month contracts with secure hardware-bound license validation. Razor-sharp math, no black boxes.
Click to claim your indicatorsBeginner portal · journal · trade tracking. Log every trade and your own observation ideas — the system replays them via an automated Sunday mirror against systematic outcomes. High-end indicators on clean Databento tick data, Supabase-backed journal, ICT-indexed reminders pulled from Twitter Spaces, lectures, and live-stream archives.
Full data + backtest engine. HOD/LOD distributions, automatic N ≥ 30 validation gating, ICT week-characteristics database dynamically mapped across historical structures. One-time $599 unlocks the engine for life.
Escape the educational discord noise. Production-ready hard-coded logic for 3 institutional models — raw Pine Script / C++ indicator source, comprehensive mathematical explainer PDF, and a developer breakdown video mapping exact entry/exit protocols.
Short-term high-probability execution
Based on liquidity sweeps, market structure shifts, and Fair Value Gaps within specifically defined kill zones.
Market Maker Buy & Sell Model
The core systematic framework mapping the complete lifecycle of institutional order flow using classified weekly regime profiles and time-weighted probability surfaces. Hard to code — that's the point.
12 Classified Profiles Integration
High-confluence rejection
High-confluence setup where a Fair Value Gap overlaps a Breaker Block, creating an exceptionally strong algorithmic rejection zone.
Each model ships with framework video, advanced protected indicator, mathematical brief, and a dedicated research channel.
Autonomous Python signal-generation agent backed by the data engine and a trade-grading system. Immediate, sub-second-latency trade entries, macro-window targeting vectors, and systemic direction triggers — restricts false signals via t-statistic validation. Every closed signal is auto-graded.
Bespoke affiliate payout layers. Leverage the 19-year quantitative narrative and marketing resources to recruit traders. Dynamic dashboards track sub-network backtest interest and reward click-through metrics in our Apple-Silicon-isolated ledger.
Apply with credentialsThree deliverables. Mathematically traceable, statistically gated, deployed on a private two-Mac-Mini cluster. For prop firms, hedge funds, family offices, and quantitative research desks.
Engineered for quantitative firms, systematic hedge funds, and family offices. Synapse provides bespoke higher-timeframe and lower-timeframe — even sub-1-minute — price action studies and modeling for every asset class we cover: Currency Futures & Forex, Index Futures (our specialty), Stocks & Options, and Crypto (BTC / SOL / ETH / XRP) — essentially CME products plus whatever the enterprise requests. Deliverables include model breakdowns, seasonal-tendency studies, and combined short-term + long-term projections built with bespoke mathematics and state-of-the-art algorithmic data modeling — all executed on private, fully isolated Apple Silicon.
Real-time, sub-minute trading signals computed on live price action — server-side inference executing on Apple Silicon MPS backend, emitting tape-validated trigger events gated by t-statistic significance. Sub-50ms compute window. Deterministic, idempotent, fully replayable against historical state.
The full clean CME Globex tick / 1-minute / hourly transaction history — at your disposal. zstd-parquet partitioned by session × weekday × risk-week regime, queryable through SQL or the Synapse Python SDK. Already validated, gap-filled, and N ≥ 30 statistically gated. No data engineering on your side.
Higher-timeframe projection combined with sub-1-minute lower-timeframe modelization — bespoke algorithmic application engineered per asset class. Each study delivers a seasonal-tendency model, microstructure regime tagging, a time-partitioned probability surface, and a p-value tested execution protocol — deployed on private Apple Silicon, fully isolated from your stack.
from synapse_enterprise import BespokeStudy, MacMiniCluster
# Two Mac Minis · sovereign · fully isolated
cluster = MacMiniCluster(nodes=2, isolation="full")
study = BespokeStudy(
assets=["NQ1!", "ES1!", "6E", "BTC-CME"],
timeframes=["1W", "1D", "1m", "30s"],
history_from="2007-01-01", # 19 yrs clean tape
gate="N >= 30 AND p < 0.05",
)
# Industry-first time-and-price classification
study.classify(method="time_x_price", window="session")
study.model_seasonal_tendency(method="bespoke")
study.project(horizons=["short_term", "long_term"])
study.deploy(cluster=cluster)
# > DEPLOYED. Mathematically traceable. No black boxes.
The Synapse engine runs continuously on private Apple Silicon against transacted CME tape flows. Every number below is computed, never estimated.
Signal generation, model training, and execution run on private Apple Silicon — orchestrated with production-grade, battle-tested components. Hover to highlight.
Signals, model training, and trade logic all run on private Apple Silicon local machines — never a shared cloud tenant. That means faster compute, complete data isolation, zero-trust verification, and a strictly zero human-in-the-loop engineering pipeline. Nothing leaves the machine that doesn't have to.
Every signal is computed locally on a private two-Mac-Mini cluster — never a rented tenant, never a shared GPU.
The 19-year CME moat is processed in an isolated local environment — no third-party data exposure.
Every signal is reproducible from raw tape — no black boxes, no discretionary overrides.
Indicator access is bound to validated hardware API keys with zero-trust handshakes.
A thoughtfully designed Python SDK that gets out of your way. Ship faster with hard-coded institutional logic and exact entry/exit protocols — all executable on your own Apple Silicon.
Raw Pine Script v6 / C++ source delivered.
N ≥ 30 validation baked into every call.
Real-time detection on transacted tape.
Math PDF + developer breakdown video.
"Synapse replaced every discretionary guess in my process. What used to be a feeling is now a number with a p-value attached."
Synapse Terminal would not exist without the work of Michael J. Huddleston.
The frameworks we engineer — every weekly profile, every kill zone, every liquidity sweep, every Fair Value Gap — sit on the shoulders of two decades of his price-action lectures, his live streams, his Twitter Spaces, and the unrelenting clarity with which he gave them away to anyone willing to listen.
We acknowledge that gratitude here — in writing, in code, and in product.
Synapse exists because you taught.
Thank you, Michael.
— The Synapse Terminal Team
Synapse runs on math, on Apple Silicon, and on the community that came before us.
That community got us here for free — through hours of public lectures, recorded Twitter Spaces, and freely shared annotations. We owe the same in return.
So we give back:
A core set of our Pine Script indicators — session typology markers, weekly bias overlays, and base kill-zone shading — ships free to every account holder. No subscription required.
Pre-market notifications for macro windows, NY AM open, London open, and end-of-week regime shifts — delivered to anyone with a free Compass account.
Claim your free seat at the Synapse Terminal. No credit card. No upsell. Walk in and use the engine.
This is how we honor the system that built the system.
The exact same products and pricing from the catalog above. Start on Compass and scale into institutional-grade model deployment. No hidden fees.
Beginner portal · journal · trade tracking
then $31.99/mo
Advanced traders · full engine
Entire Data + Backtest Engine
Institutional model deployment
MMXM model from $1,399 + $169/mo
Bespoke study · custom architecture
Bespoke quote
All tiers run on private Apple Silicon with zero-trust ingress, full data isolation, mathematical traceability, and the N ≥ 30 statistical gate.
Join operators shipping systematic, capital-preserving execution with Synapse Terminal. Start free, scale to institutional grade — all on sovereign Apple Silicon.
No credit card required for the live simulator