Our approach

One platform for data, research and execution.

Engineers in San Francisco build and run the systems that all three investment disciplines share, from the first data load to the last trade report.

Overview

We build one platform rather than three.

A credit analyst, a portfolio manager and a systematic researcher read the same data, run on the same research tools and trade through the same controls.

Engineers and data staff
58
Datasets under point-in-time control
1,200
Research jobs run per day
35,000
Based in
San Francisco, CA

Data platform

Data stored as it was known.

Market, fundamental, macroeconomic and alternative data enter through one pipeline, with the same checks and the same history rules.

Storage

Point-in-time history

Every value keeps the date it became known and every later revision, so research can ask what was knowable on any past date.

Quality

Checks on every load

Automated tests flag missing values, stale prices and unusual jumps before data reaches a model. Failed loads are held, not passed through.

Access

Entitlements and lineage

Access follows vendor licenses and information barriers. Lineage records where each field came from and every step that changed it.

An engineer reviewing code on a large monitor

Research infrastructure

Every result can be run again.

Research tools are built so that a reviewer can reproduce any result months later and get the same answer.

  • Versioned code

    Research code lives in source control and is reviewed before it is used in a decision.

  • Recorded data snapshots

    Each run stores the exact data version and parameters it used alongside its results.

  • One backtesting engine

    All teams share one simulation engine with built-in cost, financing and capacity models.

  • Managed compute

    Large studies run on scheduled cloud and on-premises capacity, with usage tracked by team and project.

Our research process

Trading systems

Execution that is measured before and after.

Every order passes through the same three stages, from the first check to the final cost review.

Execution research
  1. Pre-trade controls

    Orders are checked against client guidelines, risk limits, restricted lists and price bands before they can be sent to a market.

  2. Order management

    One order system records every instruction, allocation and change with a timestamp and the person or program responsible.

  3. Transaction cost analysis

    Each trade gets a cost estimate before execution and a measured cost after it, and desks review the gap every month.

Security

Protecting client data and systems.

Our security program is aligned with the NIST Cybersecurity Framework and with SEC requirements for safeguarding customer information under Regulation S-P. It is overseen by the Chief Technology Officer and reviewed by the Risk Committee.

Access

Staff use single sign-on with phishing-resistant multi-factor authentication. Access to client and trading systems is granted by role, reviewed quarterly and removed on the day someone leaves.

Encryption and monitoring

Client data is encrypted in transit and at rest. Systems are monitored around the clock, and alerts are triaged by an on-call security engineer.

Testing and vendors

Independent firms test our external systems each year. Vendors that hold client data are reviewed before onboarding and at least annually after.

Incidents

A written incident response plan sets roles, escalation and client notification. Plans are rehearsed with tabletop exercises, and lessons are tracked to completion.

Build the platform with us.

We hire software, data and infrastructure engineers for full-time roles, and technology interns in New York and Chicago.