Advancing Specter's AI Capabilities with SoftBank Corp.'s GPU Infrastructure

Tokyo, Japan — April 9, 2026 — We are working with SoftBank Corp. to leverage their AI data center GPU infrastructure to accelerate the development of Specter's AI capabilities, operating entirely on infrastructure based in Japan.

We will be using this infrastructure via SoftBank Corp.'s AI Foundation for Startups program, which supports startup companies in developing AI models and commercializing their businesses.

Access to scalable, high-performance compute — particularly at the level of modern GPU clusters — is a fundamental constraint in building production-grade AI systems. Through this engagement, we are able to significantly increase the speed, depth, and reliability of our model development while operating fully within Japan.

Building Core AI Capabilities

Our current focus is on developing foundational capabilities that enable AI systems to operate effectively on complex, real-world enterprise data. This involves:

• Transforming unstructured information into formats that AI systems can reliably interpret
• Preserving context, structure, and relationships within large datasets
• Creating robust pipelines that support downstream reasoning and automation
• Ensuring sensitive enterprise data is processed locally, without routing through external or overseas systems

We view this layer as a critical enabler for any meaningful enterprise AI application.

Leveraging Compute for Model Development

By leveraging SoftBank Corp.'s GPU infrastructure — we are able to:

• Fine-tune and adapt open-source models for real-world business environments
• Run large-scale experiments across different model architectures
• Rapidly iterate toward higher accuracy and reliability

This allows us to focus not just on model performance, but on building systems that can operate consistently under real-world constraints, ensuring enterprise data never leaves the country.

Toward Fully Domestic AI Systems

A core part of our vision is to build AI applications and services that are:

• Developed and deployed on infrastructure based in Japan
• Aligned with enterprise data security and residency requirements
• Continuously improved through domain-specific learning
• Operated with full auditability and enterprise-grade data security

As AI adoption accelerates, the ability to operate within a fully domestic compute environment is becoming increasingly important for enterprise use cases.

The Start of a Larger Journey

This is an early step in a broader journey to build AI-native systems designed for complex, high-stakes environments.

We are focused on building across multiple layers: from data processing to model development to application-level intelligence, with the goal of enabling more reliable and scalable AI-driven workflows.

Access to advanced compute infrastructure fundamentally expands what can be built, and how quickly.

With the support of SoftBank Corp., we are able to operate at a level of speed and ambition that would otherwise be difficult to achieve — and do so entirely within Japan, where our enterprise clients need their data and AI systems to remain.