A boutique, female-led systems engineering and R&D firm in Bromont, Québec. We integrate a business's technical and operational systems — software included — so it runs on its own terms, without depending on outside subscriptions and platforms. Agentic AI, SysML v2, and Semantic Control Theory are our methods; sovereignty, autonomy, and sustainability are the aim.
2nd & Capital inc. is a boutique, female-led engineering and R&D consultancy founded and led by Alice Figg (Co-founder & CEO), alongside Brent Figg (Co-founder & Chief Systems Architect). Incorporated in Bromont, Québec in February 2023, the firm pairs rigorous project management discipline with active R&D in agentic AI and SysML v2 — the current international standard for Model-Based Systems Engineering.
That technical work directly extends academic research co-authored during Brent's M.Sc. at Wayne State University, completed in August 2026; he is currently pursuing a PhD in Industrial Engineering there (in progress via the Global Executive Track pathway).
Most small and mid-sized businesses run on a patchwork of other companies' subscriptions, platforms, and software. We work with Canadian businesses to map their technical and operational needs as one system, and to integrate what they need so they own and control it — software included.
Agentic AI is how we work through a business's problems and needs across that whole system, so it can become self-controlled and sovereign. The same methods apply in defence and ground-systems architecture.
Rooted in Québec, working with a national and international ecosystem.
Orchestrated AI agents applied to a business's problems and needs across its systems — working agents embedded in engineering workflows, not chat interfaces.
Model-Based Systems Engineering using SysML v2, the current standard, to represent and reason about complex system architectures — from ground vehicles to organizations.
A formal framework for how meaning and control propagate through complex systems — the theoretical backbone linking our agentic-AI and MBSE work together.
“An Agentic AI Vision for Combat Vehicle Architecture Model” — presented at GVSETS (Ground Vehicle Systems Engineering and Technology Symposium, NDIA Michigan Chapter), in active collaboration with Gregory Pappas and Dr. Stephen Rapp of the U.S. Army DEVCOM Ground Vehicle Systems Center (GVSC/TACOM).
This work sits within Québec's broader defence-industrial and advanced-manufacturing ecosystem — the domestic capability our research is designed to serve.
Client engagements are confidential by default; we do not publish client or partner names.
Our research programme studies how complex technical and organizational systems can be composed, monitored, and verified using formal, open-standard models. Current themes: composable system architectures; sequential decision-making under uncertainty; and Bayesian monitoring of governance and compliance data. Working papers are listed here as they are released.
PRAXIS: Parametric Recursive Architecture for Transdisciplinary X-domain Integration in Systems — Brent Figg, Wayne State University (Industrial and Systems Engineering).
Proposes a minimal formal kernel — a belief-state model built on a partially observable Markov decision process — from which a defined class of complex systems can be expressed as parameterized instantiations rather than designed from scratch. The paper constructs two boundary cases and a composition operator showing how separately built systems can be combined.
© 2026 Brent Figg. Free to read, download, and cite with attribution; all other rights reserved. Full text (PDF, 19 pages)
TEMPO: Time-Indexed Evidence-Based Marginal Gains under Partial Observability — Brent Figg, Wayne State University (Industrial and Systems Engineering).
Proposes an architecture for compounding marginal gains where time passes, resources are finite, interventions interact, and the value of action is only partially observable. The central idea is that opportunity cost can be formalized as a time-indexed component of a partially observable Markov decision process's reward, with latent opportunity windows modelled as a hidden Markov model. The framework combines Gaussian-process Bayesian optimization, a Markov decision process formulation, and an existing structured decision-analysis interface (VADM/LDW) as a governance layer. High-performance sport development is the illustrative example; the paper claims no empirical results.
© 2026 Brent Figg. Free to read, download, and cite with attribution; all other rights reserved. Full text (PDF, 33 pages)
SENTINEL: Structured Evidence-Normalized Transparency and Intelligence for Non-profit Evaluation in Landscapes of Threat — Brent Figg, Wayne State University (Industrial and Systems Engineering).
Examines how governance quality in small and mid-sized non-profit organizations could be monitored systematically from public regulatory filings, and why existing audit mechanisms struggle to do so at population scale. The paper proposes a probabilistic monitoring architecture and a synthetic validation design, and analyzes the approach's failure modes and limits.
Scope: a theoretical working paper. It names no real non-profit organizations, makes no claims about any existing non-profit, and proposes validation on simulated data only; no empirical results on real organizations are claimed.
© 2026 Brent Figg. Free to read, download, and cite with attribution; all other rights reserved. Full text (PDF, 29 pages)