What we do
Discovering Signal With Machine Intelligence.
Our business model is simple: give exceptional researchers and engineers the data, compute and capital to build AI-focused trading models, then hold those models to the highest standard of evidence.
Our approach
Putting Ideas Into Production.
A Balboa Line model team is small, cross-functional and fully accountable. It owns a hypothesis from the first experiment to the live order flow, supported by a technology stack designed to shorten every step in between.
Our firm is purpose-built to support this loop. Shared infrastructure handles market access, data, compute, risk and compliance, so teams spend their time on what differentiates them: the models.
A Platform for Performance
Ongoing investment in engineering talent across market access, data, compute, research infrastructure, compliance and business support, so researchers can focus on research.
Small Teams, Shared Scale
Teams of three to a dozen people run their own models with their own capital allocation, while drawing on the scale, data and controls of the whole firm.
AI-Native From Day One
We did not bolt machine learning onto an existing business. Our data, tooling and governance were designed for learning systems from the first line of code.
Inside our model teams
Many Models. One Line.
No two teams look alike. Some trade a single product at very short horizons, others run cross-asset portfolios over days and weeks. All of them build their own models and are measured on live results.
Where our models trade
- Strategies across many asset classes, or focused on a single product
- Horizons from microseconds to weeks, at any trading frequency
- Teams from three people to a dozen, each with its own capital allocation
- 30+ venues across the Americas, Europe and Asia
- Equities, futures, FX, listed options, rates and digital assets
Evidence over opinion
Every Model Has an Owner, a Card and an Off Switch.
Machine learning in markets is only as good as the controls around it. Ours are designed in, not added later.
Model Cards
Every model documents what it predicts, which features it uses, how it behaves at the edges of its data and what will cause it to be retired.
Independent Review
No model trades until it has been reviewed by researchers, risk and compliance staff who had no part in building it.
Pre-Trade Limits
Position, notional, rate and loss limits are enforced outside the model in infrastructure the model cannot modify.
Kill Switches
Every system can be halted in milliseconds by its owner, by risk, or automatically when behaviour drifts from its card.
Qualities we prize
What Our Best Researchers Have in Common.
Rigor
Embracing evidence over intuition, and holding ideas to the same standard whether they come from an intern or the CEO.
Clear Communication
With teammates, core engineers and management, so that decisions are made on shared understanding.
Adaptability
Markets change suddenly and models decay. We prize people who notice early and adjust without drama.
Calm Under Pressure
Staying steady when positions, systems or data behave unexpectedly, and making the next decision well.
Embrace your role
How Model Teams Are Built.
Team structures vary, but these are the roles you will find on most of them.
Quantitative Researcher
Uses our simulators and historical data to find and validate patterns, designs features and defines the models that express them.
Machine Learning Engineer
Builds the training pipelines, serving code and evaluation tooling that turn a research result into a system that trades in production.
Trading Operations
Monitors performance and risk in real time, produces the reporting that drives model review, and leads strategic projects for the team.
World-class engineering
The Feedback Loop Between Models and Machines.
Model teams and core engineering sit side by side. Engineers build the low-latency connectivity, data infrastructure and risk frameworks every model depends on, and researchers feed back what the next model needs.
Culture
A Culture of Curiosity.
A results-oriented environment during the trading day, and a generous one outside it: team outings, research talks and causes we support together. Take the next step toward joining a team as distinct as the models we build.
Reach your greatest potential
Help Us Build the Next Market.
We bring the platform, the capital and the research culture. You bring the ideas. Take the next step to learn about our teams, our hiring process and life at Balboa Line.