SA Hub
Choose a starting point
Explore example strategies and shared research, or bring your own ideas to the workspace.
Browse researchInvitation-only beta
Research and trading infrastructure for small, AI-native teams.
Choose a strategy, test it against market data, and review the evidence with your team.
Experiment record
- Strategy
- Example strategy (v1)
- Market
- MES
- Data window
- Aug 3–23, 2026
- Evaluation window
- Aug 17–23, 2026
- Assumptions
- Historical bars. Separate evaluation window.
- Limitations
- Historical results do not establish future performance.
Lab
Test against market data
Run historical backtests with clear assumptions, time windows and recorded configurations.
Explore the labDesk
Review the evidence
Compare results, discuss the next experiment and keep operational context with your team.
See how teams workThe workflow today
Make every experiment useful.
Keep the question, configuration and evidence together. Learn from completed tests and failed experiments, then decide what to try next.
Select
Start with a strategy and an instrument. Set the question and the configuration you want to test.
Test
Submit a supported backtest in Lab. Follow its status and review the data window and assumptions.
Review
Compare completed results, inspect limitations and keep an unsuccessful experiment in the record.
Iterate
Use what you learned to choose the next test. Strategy tuning and model training are different workflows.
Inside Lab
A question. A configuration. A record.
See the current backtest setup. Every experiment starts with an explicit data window and assumptions.
Hosted access is by invitation. Tools depend on your team’s permissions and available data. Customer model training is not available in the hosted app.
Who does what
Your research and decisions. Our infrastructure.
Your team decides what to investigate and what the evidence means.
Your team
The research questions, strategy choices and decisions about what to test next.
115
The hosted software, team access boundaries and infrastructure that keep the workflow connected.
In development
A research loop that keeps learning.
Our direction is a weekly cycle: refresh the data, train a candidate, evaluate it after costs and risk, then promote it or keep the current model. Hosted customer training is not available today.
A weekly research rhythm
- 01 / Refresh
New week. New evidence.
Add the latest data. Keep a separate evaluation window.
- 02 / Train
Challenge the current model.
Train a candidate with a recorded dataset and configuration.
- 03 / Evaluate
Make improvement earn its place.
Compare against the current model after costs and risk checks.
- 04 / Decide
Promote. Or keep what works.
Review the evidence. Keep the current model if the candidate falls short.
Our focus
Small teams. Serious tools.
115 is a hosted platform and lab for small, AI-native teams. We develop it alongside the daily work of our own research desk, then use that feedback to improve the tools available to invited teams.
Meet 115Risk disclosure
- Trading involves substantial risk of loss and is not suitable for everyone.
- Hypothetical and backtested results have inherent limitations: they are prepared with the benefit of hindsight and do not represent actual trading.
- Past performance does not guarantee future results.
- Nothing on this site is investment advice, or a recommendation to buy or sell any security, futures contract or other instrument.
For your team
Start with your research workflow.
Tell us what your team wants to test and how you work today. We will walk through the hosted platform and the access available for your team.
Request beta access- Tell us about your team.
- Walk through the platform with a person at 115.
- Agree your team’s access and onboarding.