Research-ready data
Access normalized market, fundamental and reference data through Data Gateway APIs and Python workflows, with point-in-time context, consistent instrument identities and visible sources.
Quant research
From hypothesis to evidence you can trust.
Data Gateway, robust backtesting and reproducible analysis around your models - from research-ready inputs to a strategy you can challenge and a portfolio you can evaluate.
- Data
- Backtesting
- Research
- Portfolio Analytics
For quantitative researchers, systematic investment teams and research engineers.
The operating reality
The result has to survive the next run — and the next researcher.
A promising backtest is the beginning of the discussion. Was that fundamental value available at the time? Did the universe include delisted instruments? Which costs were assumed? Can a colleague recover the inputs and reproduce the result after a provider revises its history?
When those answers depend on one notebook or one researcher’s memory, review becomes reconstruction. Research infrastructure should keep the data, assumptions, configuration and results close enough that the team can challenge an idea and decide what to investigate next.
Can another researcher rerun this result, challenge its assumptions and assess its portfolio impact?
Built for your team
Build research on inputs and results you can trust.
Robust backtesting
Test strategies in Backtester with explicit costs, slippage, borrow and financing. Challenge results with walk-forward analysis, parameter stability and stress periods.
Reproducibility
Keep inputs, assumptions, configuration and outputs with every run so another researcher can recover the result and challenge its assumptions.
From strategy to portfolio
Evaluate accepted signals against portfolio weights, exposures, liquidity and risk before turning a research result into a proposed allocation.
A workflow in practice
Take a hypothesis through a repeatable research process.
An illustrative workflow: a researcher tests a strategy, challenges the result and evaluates a candidate allocation. Each stage has an explicit input and an output the next person can inspect.
- 01
Establish the inputs
Select sources, instruments and observation dates. Check the available history and point-in-time coverage.
Data Gateway - 02
Run the experiment
Make strategy parameters and execution assumptions explicit, and retain the run configuration.
Backtester - 03
Challenge the result
Review costs, walk-forward behaviour and sensitivity before accepting the evidence.
Backtester - 04
Evaluate portfolio fit
Assess a candidate allocation against weights, exposures and risk before proposing a portfolio change.
OneBook
A completed run is the starting point for review.
Illustrative product view · View full size ↗Inspect strategy metrics and charts, then challenge the assumptions and execution settings behind the result.
How it fits your stack
Keep your models. Connect the research infrastructure.
Use Data Gateway through APIs, Python SDKs and permissioned MCP alongside your existing notebooks, datasets and research tools.
Run the Apache 2.0 Backtester locally, on-premise or in your cloud. Add portfolio and risk workflows when accepted research is ready for a book.
Use directly
Start with the open-source Backtester or a Data Gateway integration. Choose the research workflow first, then add managed capabilities where they are useful.
Integrate selectively
Keep your Python models and notebooks. Use APIs, SDKs and permissioned MCP for supported data access, and connect portfolio analysis when a strategy reaches that stage.
Deploy privately
Run the open-source engine locally or in your environment. Confirm managed data, PRO tooling, licensing and private deployment requirements separately.
Can we keep our notebooks and existing models?
Yes. Data Gateway APIs and SDKs can sit alongside your existing research code. Adapting a strategy to the Backtester requires mapping its inputs, signals and execution assumptions to the engine’s supported interfaces.
Is all data point-in-time and free of survivorship bias?
Those properties depend on the dataset, provider and how a research universe is constructed. We confirm available timestamps, revision history and instrument coverage for the requested sources. Using a common API does not by itself remove look-ahead or survivorship bias.
What is open source, and what is part of PRO?
The Backtester engine is available under Apache 2.0. PRO adds managed market and fundamental data, advanced research workflows and hosted or private tooling. Dataset rights, entitlements and deployment scope are confirmed separately.
How do we evaluate it on our research process?
Choose one strategy and specify the universe, sample period, data requirements and execution assumptions. Walk through the inputs, a reproducible run and validation questions before considering a broader integration.
What does moving from our current setup involve?
Start with one workflow, its source systems and the data it needs. We agree mappings, responsibilities and evaluation criteria before introducing live data. Existing systems can remain authoritative while the new workflow is validated; any migration is scoped separately.
How is pricing structured?
Pricing depends on the selected products, data entitlements, integration scope and deployment. The open-source Backtester engine is available under Apache 2.0; managed data, PRO tooling and private deployments are evaluated separately. We establish the requirements before proposing a commercial scope.
Start the conversation
Tell us about your research process.
Tell us where research becomes infrastructure work: sourcing data, reproducing a run, validating a strategy or evaluating portfolio fit. Bring one representative workflow and its assumptions. We will discuss how it could connect to the tools and models you already use.
Start with a description of the workflow. Please do not submit confidential holdings, client information or access credentials.
Or explore the open-source BacktesterWhat would you like to explore?
Required fields are marked with an asterisk.
Thank you for getting in touch.
We will review your workflow and reply to discuss the next step.
Return to QuantJourney