The Agentics Foundation brought together a cross‑section of Canada’s emerging agentic ecosystem AI researchers advancing planning and reasoning methods, developers shipping vertical‑specific agents, founders testing agent‑native product surfaces, and infrastructure teams focused on reliability, observability, and execution guarantees. Across discussions, a consistent set of engineering gaps surfaced: reproducibility across runs, deterministic tool invocation, multi‑step evaluation, and the absence of shared standards for agent behavior.
Railtracks joined as an open‑source framework partner built for structured, reliable, and extensible agent infrastructure that developers can deploy in real‑world environments. This collaboration marks the beginning of a broader Vancouver, BC builder network committed to publishing open reliability patterns, developing shared evaluation benchmarks, running workshops and hackathons on agent architecture, supporting teams adopting the framework, and contributing to emerging standards for agentic systems.
