Prove the edge is real.
quantOS counts every test, files the memo with its sources, and keeps the trail, so your committee and allocators see what was tried, what was killed, and why.
Send us one backtest, get a counted-trial verdict. Paid pilots for emerging funds, by application.
Or run the open-source Python library locally
01
Hypothesis
Kill condition written
02
Trials
Every variation counted
03
Memo
Sources cited
04
Trail
Timestamped for review
- 01 · Hypothesis Board
- Every claim has a kill condition
- Killed ideas stay on the record so the desk never quietly repeats a failed backtest.
- 02 · Counted Trial Ledger
- Every variation is logged
- Run sweeps in Python, or log a QuantConnect or VectorBT result on the same counter.
- 03 · Cited Committee Memos
- Ready for the meeting in PDF or Word
- Deliver links the thesis, the dataset fingerprint, and the surviving study in one packet.
Shared Success · How It Works
From first idea to committee memo.
Your OMS runs trades. quantOS governs the research before the trade—so every test, memo, and decision lives in one calm workspace.
Log the idea
Nothing gets lost in notebooks, and killed ideas aren’t quietly re-run next quarter.
Every idea starts with what would prove it wrong. Killed ideas stay on the board so the team does not re-run them.
Prior-work check activeInventory revision spread
Dies if post-filing drift reverses within two quarters.
Supplier lead-time residual
Dies if signal decays under execution delay.
Overnight gap reversion
Killed: the opening auction explained the entire result.
Semiconductor capex residual
Cleared: residual holds with sector beta hedged.
Test it
You can show your committee how much you tried, not just the one run that worked.
Every test is logged as you work—whether run in quantOS, from Python, or imported from QuantConnect and VectorBT.
9 trials countedSemiconductor capex residual
quantOS Studio
Supplier lead-time parameter sweep
AlphaEngine CLI
Historical parameter grid import
VectorBT Import
Opening gap mean-reversion study
QuantConnect Import
Decide
Walk into the meeting with a finished packet instead of rebuilding history the night before.
Semiconductor Capex Revision Sleeve
Prepared for the Investment Committee with the hypothesis, the counted trial history, and the dataset version attached.
Kill condition
Cleared out of sample
Trials counted
9 runs on the ledger
Mandate gates
All desk limits cleared
Prove it
When an allocator asks how you got here, you send the record.
A single timeline of what the terminal sealed, what the portal ran, and what the desk decided.
Shareable trailCommittee memo filed
Semiconductor Capex Sleeve · PDF & Word
Study sealed from terminal
Validated against point-in-time dataset v3
Book overlap check
Compared against active Equity Pod holdings
Hypothesis retired
Overnight gap thesis kept on permanent record
Platform Overview
One record from idea to decision.
Keep your existing backtester. quantOS connects your hypotheses, trials, datasets, and committee memos into one continuous story.
- 01
Hypothesis
Start with a clear claim and the condition that would kill it.
- 02
Counted trials
Log every test across Python, QuantConnect, VectorBT, and the web.
- 03
Committee memo
Generate a clean PDF or Word packet with every source cited.
- 04
Versioned trail
Share a timestamped research record with your committee or allocators.
Explore Workflows
See how a desk runs on quantOS.
The model picks the tools. Deterministic Python computes every number.
File the memo
Move from a falsifiable hypothesis to a cited committee packet in PDF or Word.
Version the data
Lock point-in-time constituent universes and retain delisted names before testing.
Seal locally
Run parameter sweeps on your own machine and sync sealed studies to your team.
Ask the floor
Run eight guided research checks or connect your assistant over MCP.
Open-Source Foundation
Run AlphaEngine on your machine.
Anyone can install the open-source Python library to run parameter sweeps and statistical checks offline. When a strategy is noise, it says so.
$ alphaengine demo verdict likely_noise A crossover on a random walk has no edge, and the verdict says so rather than flattering it. That is this working, not failing.
Data Custody
Keep your math and data yours.
The AlphaEngine library and its math run offline on your machine. Login, workflows, reporting, and sealing use the network, and full return series are sent only when you pass --share-series.
Explore Solutions · Who It Is For
Built for every seat at the table.
Designed for emerging hedge funds and systematic pods managing up to $1B AUM.
- PM & CIO
Decide from one cited packet
Walk into the committee meeting with the hypothesis, the counted trial history, and the risk checks already compiled in PDF or Word.
- Quant Researcher
Keep your backtester, keep the count
Run tests in Python, or log a QuantConnect or VectorBT result. The counter keeps what was tried.
- Chief Compliance Officer
Answer the committee with the record
When an allocator asks how a strategy was approved, share the versioned trail from the first idea to the committee decision.
For Allocators
Let your emerging managers share a versioned research trail with you.
See the hypotheses tested, the ideas killed along the way, and the cited committee memo behind a manager’s process.
Simple Packaging
Flat team pricing, no seats.
One flat monthly rate for the whole research pod, with separate self-serve tools for individual researchers and a free open-source Python library.
Get Started
Start from the record.
Bring your research pod onto quantOS, or run AlphaEngine locally in Python.