Introduction

Xavi.app is building Northern Canada’s first sovereign AI-focused GPU cloud, headquartered in Prince George, BC.
Our mission is simple: deliver enterprise-grade compute at hobby-grade prices—without sending Canadian data across International Borders or into the hands of Foreign Adversaries.

Learn More

Bottom Line

Transforms Northern BC’s 98% renewable power into an up-cycled, carbon-light GPU cloud that reaches positive cashflow on its first GPU Cluster, compounds capacity every 12-18 months, and delivers 40 %+ EBITDA—while giving Canadian governments, hospitals, and innovators a sovereign,CLOUD-Act-proof AI backbone at up to 80 % below hyperscaler pricing.

Learn More
 

Private Investors

Turning Cheap GPUs + AI Tools into High Returns

Bargain GPU Strategy

  • Used NVIDIA A100: ~$6,500 CAD.
  • 8 V100 server: $10,000 CAD (~$1,250/card).
  • Tested, cleaned, deployed in Prince George data centre.

Fast Returns

  • A100 rents at ~$2/hour; 60% use earns ~$10.5k/year.
  • 8-V100 server earns ~$30k/year.
  • Post-purchase, profits fund more GPUs.

AI Services & Market Edge

  • AI tools (chatbots, summaries, image generators).
  • Canadian market lacks affordable AI.

Cost, Green & Exit Paths

  • 50-80% below AWS; <5ms latency in Western Canada.
  • 98% clean hydro, reused hardware cuts e-waste.
  • Attractive to green tech firms, telcos, data centres.

Public-Sector Partners

Saving Money & Securing Data with Local GPUs

Data Security & Compliance

  • Prince George data centre keeps data in Canada.
  • Meets PIPEDA, BC FOIPPA, “Protected B” standards.
  • Built for ISO 27001, SOC 2 Type II.

Cost Efficiency

  • A100 at ~$2/hour, 50-80% less than AWS.
  • One-click AI tools: chatbots, translation, analytics.

Local & Green Impact

  • Jobs in AI, cybersecurity, data ops in Northern BC.
  • Reused hardware (~80% less e-waste).

Sovereign AI Strategy

  • Aligns with Ottawa’s $2B AI Compute Strategy.
  • Builds Canada’s independent AI infrastructure.

Why Now?

Global Demand Surge
Generative AI and large language models are driving unprecedented GPU demand. Canada imports over 80% of its AI compute from foreign hyperscalers, with domestic needs exceeding supply by 10×.
Policy Support
The $2B Sovereign AI Compute Strategy and provincial initiatives prioritize local infrastructure, offering grants and incentives for clean tech and data sovereignty.
Market Timing
Canada’s AI ecosystem is maturing, with SMEs, universities, and public sectors seeking affordable, local compute solutions to compete globally.
Cost Arbitrage
A global surplus of used GPUs (e.g., NVIDIA A100 at ~$6,500 CAD vs. $13,000 new) enables 30–70% savings, lowering capital barriers for rapid scaling.
Renewable Power Edge
Prince George’s 98% hydro power at ~$0.06/kWh provides a 50% cost advantage over fossil-based regions and aligns with ESG goals.
Competitive Advantage
Low-latency (<5ms) compute in Western Canada, paired with 50–80% cost savings vs. AWS, positions us to capture underserved markets.

Desired Roadmap

NVIDIA 40GB A100 Tensor Core GPUs ($4,000–$6,500 CAD each)
Q4 2025 20 GPUs (PoC pod) ▼
Milestones: Beta with partners and 2 local SMEs
Funding Mix: CFDC loan + founder equity
Q2 2026 100 GPUs ▼
Milestones: Public launch, break-even at 65% utilisation
Funding Mix: NDIT grant + revenue
Q2 2027 300 GPUs ▼
Milestones: Second pod, onboarding first gov’t workloads
Funding Mix: BDC tech loan + angel round
2028+ 1,000+ GPUs ▼
Milestones: Cross-province edge sites, federated fine-tuning service
Funding Mix: Series A / strategic partner

What We’re Building

Flexible GPU Clusters
We combine powerful consumer cards (RTX 3090, 4090, 5090) with enterprise GPUs (A100, H100, V100) to create custom AI compute pools. Customers can choose a single cluster or split resources across multiple smaller units—whatever fits their project.
Easy Self-Serve Portal
A simple web dashboard and command-line tools let you pick your GPU type, set up virtual machines, containers, or bare-metal servers, and start training or running models in minutes.
Compliance-First Design
All data stays in Canadian facilities. Optional air-gapped nodes and strict security controls ensure you meet any government or healthcare rules.
Green & Circular
We reuse and refurbish hardware, capture waste heat for our buildings, and run on clean hydro power—keeping costs low and emissions near zero.
Pick Your Price & Provider
Our platform links to multiple GPU clusters. You see who’s offering what rates and can rent directly from the provider that works best for you.

How We Win

  • Price Leadership: Upcycled hardware and free air cooling yield 50–80% lower GPU-hour rates than AWS (ca-central-1 region). Our customers access world-class compute at a fraction of typical cloud costs.
  • Western Canada Latency: With local clusters, users see sub-5ms latency in Western Canada—versus 40–80ms cross-border, which is crucial for interactive AI and data-sensitive workloads.
  • Sovereign & Regulatory Assurance: Our 100% Canadian-owned operation insulates clients from extraterritorial laws (like the CLOUD Act) and meets strict compliance requirements.
  • Hands-On Support: Unlike hyperscalers, we offer live Matrix chat and direct phone access to experienced engineers—real people, not chatbots.
 

Investor Contact:

Required
Required