Pre-alpha · research preview

A personalized
trading harness.

Software that wraps one human trader in a full engineering stack — so a single trader, or a multi-node desk, can run a real book around the clock.

modelhuman-in-the-loop
capitalown book
runtime24 / 7
stackpy · rust · ts

native to Optimus · OpenClaw · Hermes · Claude Code · Codex⌘K

ES·NQ·YM·RTY·CL·GC·SI·ZB·ZN·6E·6J·BTC·ETH·SOL·ES·NQ·YM·RTY·CL·GC·SI·ZB·ZN·6E·6J·BTC·ETH·SOL·
the desk, live

The whole pipeline is software we own.

From signal to fill — an engine that runs the desk continuously and keeps a human in the loop on the calls that matter.

115 // deskLIVE
enginelive
agents5 active
riskwithin limits
data feedstreaming
backtestqueued
signalillustrative
agent:researchwalk-forward · MNQ 2019–2025
enginedeterministic replay · state ok
riskexposure check · within limits
databars streaming · continuous
agent:opsreconcile · clean
engineering

Built like a systems company.

115 is software first — a production platform we design and operate ourselves.

Real-time execution engine

An event-driven core with deterministic state we can replay and audit down to the fill.

A time machine for strategies

A walk-forward lab: optimize on one window, prove out on the next, rolling across years and regimes.

Self-optimizing modelsbeta

Models that re-optimize every session — recursive improvement, applied to a live book, with a human in the loop.

Agentic operations

Specialized agents research, backtest, and help run the desk — where it counts, a person makes the call.

Data we build ourselves

Datasets and real-time sentiment signals we engineer in-house. Signal that has to be built, not bought.

Risk & reliability

Position limits, drawdown controls, and kill-switches live in the core — not bolted on.

In pre-alpha. A research preview.

One trader, amplified into a desk.

We don’t take outside capital and we don’t manage money for clients — we trade our own book. That independence lets us move fast and stay honest about what works.

See how it’s built