NEWSROOM · ANNOUNCEMENTS & INSTITUTIONAL MILESTONES
Published announcements from Reinventy open the page — each one readable in full. The institutional milestones below mark Reinventy's development as an enterprise. Each entry is dated to the quarter, categorized, and traced to a publicly citable source. Further information on any item is released under partnership.
ANNOUNCEMENTS
Reinventy outlines its approach to sovereign edge autonomy: the Tin Man cognitive cluster on NVIDIA Blackwell, the air-gapped Reinventy Shield deployment platform, and AI-driven on-prem materials discovery — deterministic, locally controlled, cloud-independent.
Read more
MILESTONES
Ordered by institutional category: corporate formation, ecosystem positioning, intellectual property foundation, and technical achievement.
Q3 2025 · CORPORATE
Reinventy Solutions Corp. is incorporated under the Business Corporations Act of British Columbia (BC1513581). The corporate seat is Victoria, British Columbia. Incorporation marks the formal institutional structure under which the company's foundational technology development and partnership engagements proceed.
Source: British Columbia Corporate Registry (publicly verifiable; BC1513581)
Q1 2026 · ECOSYSTEM
Reinventy Solutions Corp. has been admitted to NVIDIA Inception, NVIDIA's partner program for technology developers building on the NVIDIA platform. The membership supports Reinventy's development of edge-deployed cognitive architectures running on NVIDIA Blackwell embedded hardware, including the 12-core Tin Man cognitive cluster.
Source: NVIDIA Inception membership · cited in Sprint 2 Performance Report header
Q1 2026 · INTELLECTUAL PROPERTY
Reinventy has deposited foundational patent priorities at the Canadian Intellectual Property Office. The filings span the technology layers underpinning the company's composite platforms — advanced materials, electric propulsion, and cognitive control architectures. Application metadata, claim language, and jurisdictional extension strategy are reserved for partnership-level disclosure.
Source: Reinventy IP filings · generic posture per CONTENT_STRATEGY §5.3
Q2 2026 · TECHNICAL
Reinventy has closed Sprint 2 of the Tin Man cognitive cluster development on NVIDIA Jetson AGX Thor 5000, the first publicly documented Nemotron-3-Nano-Omni V3 edge deployment running 100% NVFP4 native — including the visual and audio encoders that the official NVIDIA reference recipe leaves in BF16. The Sprint 2 report is externally distributable under partnership.
Source: Sprint 2 Performance Report EN.md · Executive Summary
Q2 2026 · TECHNICAL
Reinventy Solutions has closed Sprint 3 of the Tin Man cognitive-cluster development program. The deliverable extends the cluster from a single-core multimodal cognition baseline (Sprint 2) to a 14-core voice-interactive operational platform: voice pipeline production-grade (speech recognition English baseline, multilingual text-to-speech streaming, language-model orchestration, audit-trail observer); low-latency voice loop end-to-end; Odino inference engine extended with streaming token delivery; new cognitive cores activated (text reasoning, visual perception cascade, router with voice-intent classification, audit observer); conscience layer (meta-cognitive observation) with full event coverage validated and persistent across container lifecycle. Sprint 4 introduces direct motor and actuator control architecture for autonomous platforms with the flight controller serving as dormant safety pilot, and continues enhancement of the Odino inference engine. Capability briefs and integration discussions are released under non-disclosure agreement.
Source: Tin Man v2.2 Sprint 3 closure (2026-05-17) · capability brief under partnership
Q2 2026 · TECHNICAL
Reinventy Solutions has released Tin Man v3.0, a major release migrating the cognitive cluster to NVIDIA JetPack 7.2 with every inference engine rebuilt on-device. v3.0 establishes a unified native-NVFP4 runtime spanning vision, audio, and language on a single NVIDIA Jetson AGX Thor — a complete sense–think–speak system, air-gapped and fully on-device. Measured gains over the previous generation: ~+36% sustained LLM throughput, up to −24% voice first-audio latency, and ~−63% object-detection latency. Capability briefs and the v3.0 release report are released under partnership.
Source: Tin Man v3.0 release (June 2026) · capability brief under partnership
Q2 2026 · TECHNICAL
Development of the Tin Man cognitive cluster has continued beyond the v3.0 major release with two incremental releases, v3.2 and v3.3. The release line consolidates the unified native-NVFP4 runtime established in v3.0 on NVIDIA Jetson AGX Thor — deepening validation coverage and hardening day-to-day operation of the cluster. Quantified results follow the company's claims discipline: figures are published only once measured and sourced, and are released through capability briefs under partnership.
Source: Tin Man v3.2 / v3.3 development record · capability briefs under partnership
MORE UNDER PARTNERSHIP
Each milestone above is the L0 PUBLIC declaration. Substantive details — including the externally distributable Sprint 2 report, the IP filing roster, and the operational roadmap — are released under partnership. Engage to begin the conversation.
Direct: engage@reinventy-solutions.ca