Defensible, documented and reproducible analysis for federal, provincial and municipal teams, delivered end to end by a senior quantitative specialist. Fixed price, typically under $40K, in English or French. You own the code, the data and the deliverables.
Quantitative work for programs that need numbers they can defend: in a briefing, in a report, or in front of a review committee.
Price and market modelling, comparative analysis of pricing systems, household and industry impacts, offset credits.
Forecasts of demand, costs and emissions, scenario analysis, and independent verification of existing models.
Cost and liability estimation, evaluation of projects and claims, and risk analysis with stated assumptions and ranges.
Reference datasets, predictive models, model evaluation and uncertainty quantification.
Scientific graphics, dashboards, interactive tools and automated data pipelines that refresh from source.
Scoped work packages as a subcontractor: modelling, forecasting, machine learning and data work, delivered under the prime's process.
Each engagement starts from your question and your data, and ends with something your team can use and rerun.
Every request is confirmed in writing within one business day: the question, the data, the deliverable, the deadline.
All processing is done in code, so every number can be regenerated from the source data, by us or by your team.
Predictive models are tested on data they have never seen, and the uncertainty is stated, not hidden.
A methods note and the code ship with every deliverable. Every published figure is checked twice.
A Canadian corporation based in Montréal.
Processed and stored in Canada, or in your environment.
Work and deliverables in English or French.
Code, data and documentation are yours.
Kalman filters, particle filters and volatility models running in production every day, validated out of sample. The same methods forecast demand, emissions, prices and costs.
Built single-handedly: 90+ dashboards, real-time data infrastructure and automated reporting.
At QuantM Alpha: statistical models, an execution engine and automated performance and risk reporting for institutional investors.
On a sample document set, the engine flagged 4 of 5 management claims contradicted by the financial statements, each cited to the source page.

I founded Dark Matter R&D, a Canadian company based in Montréal, to bring production-grade quantitative modelling to organizations that need numbers they can defend. I build forecasting models, data pipelines and analytics systems that run every day, and previously built quantitative systems for a $41M portfolio at QuantM Alpha.
For public-sector teams, that means analysis that holds up: every figure traceable to its source, every model validated before anyone relies on it, and code your analysts can rerun after handover. You work directly with me, in English or French, from the first call to the final deliverable.
Yes. Most engagements are scoped at a fixed price that fits a simple contract under $40K including taxes, with milestone payments. Larger work can go through a competitive process or through a prime contractor.
Yes. We take scoped work packages under a prime contractor: modelling, forecasting, machine learning and data work, delivered under the prime's process.
It is processed and stored in Canada, or stays in your environment if you prefer. Confidentiality terms are accepted as standard, and nothing is shared or reused without your permission.
Yes. Company details are above, and the capability statement is available in English or French on request: ryan@darkmatter.financial.
Send the question, the data you have and the deadline. You get a written scope and a fixed price, usually within 48 hours.
If an existing tool or another team already does this well, I'll tell you that instead.