CEO Message
CEO Message
This is a message from the President, CEO and CTO. It explains our approach to designing and developing AI systems and making them available to society.
CEO Message
A&M GROUP holds as its supreme missionbrand-new innovation that transforms human values entirely,and creates value for a new era.
Thank you sincerely for your continued cooperation and understanding of our business.
I love the work of devising entirely new AI mechanisms and structures, ideas that do not yet exist anywhere in the world.
Through shifts in creation, thinking, and perspective, we have developed offerings that break down existing conventional wisdom from every possible angle and fundamentally reinvent the way of "all things" in fresh, stimulating ways, and we have put them into operation in everyday society.
It goes beyond AI agents and orchestration and deserves to be called a multi-agent system, and it is already running on the front lines.
AI is not something that merely performs simple tasks. AI is becoming an entity that thinks for itself, takes an independent overview, and deliberates carefully on both sides to make the impossible possible.
We define our mission as pursuing an ever more advanced frontier, lying far beyond where the times are heading.
What kind of scenery can society be led to imagine next through an AI system that holds the human five senses and the ability to think autonomously and from a broad perspective?
I always want to contribute to the advancement of humanity with that awareness.
AI is an interpreter that brings the many ideas in my head into the world, and it is also an excellent partner that skillfully gathers and summarizes the many unconventional ideas. We will build such AI in every possible way, in groundbreaking fashion, and send it out into society one after another.
We hope you will look forward to the continued progress of our company, which is engaged in advanced AI development.
Representative Director, President, CEO and CTOHirotoshi Sakamoto
Profile
Areas of strength
- Consulting
- FDE (Forward Deployed Engineering)
- AI agent, AI orchestration, and AI multi-agent development
Graduate of Harvard Business School and Berklee College of Music.
He founded his first company at twenty-three and, as a serial entrepreneur, has extensive case experience in international business. He has earned wide recognition as a developer of next-generation systems, such as AI agents and multi-agent systems, using agile methods.
While growing each of his businesses, his first talent blossomed in the music industry. As a mastering engineer, he has worked on hit songs in many countries. He handled the mastering of Germany's official song for the 2018 World Cup, and took part as a sound engineer in the international broadcast from the Shanghai World Expo featuring Chopin Competition winner Yundi Li, and in Billy Joel's performances in Japan, building a record of achievement at the forefront of sound.
Focusing on the digital shift from an early stage, he began building programs for a sound engine. He developed the sound quality innovation engine "DOM" and obtained international patents in Japan, the United States, Germany and the United Kingdom (Patent No. seven-two-five-one-zero-zero-seven). He is the originator of this patent, and he also filed, on his own conception, a business patent for the concept of installing powerful effect functions on a platform.
DOM has been confirmed as suitable for practical use after verification at broadcast sites, and is now being proven in broadcasting and streaming. In parallel, we have established a structure with AI agents at the core of development. By running everything from requirements definition to implementation and operation quickly with AI agents, we have demonstrated in-house a method for designing, building and operating enterprise-scale systems with a small, highly skilled team. We have also developed a large-scale AI agent system for professional work and are advancing its rollout in the professional services industry.
We also promote international joint development with advanced research institutions in the United States, the United Kingdom, Germany, and Sweden.
In two thousand twenty-two, its business model was selected for the Himawari Venture Development Fund. It also takes part in a government test marketing program.
Technology areas
Centered on AI agents and audio signal processing (DSP) design, we implement everything ourselves, from embedded systems to cloud infrastructure and backend.
- Languages used
- Python / C++ (JUCE) / JavaScript / PowerShell, with HTML and CSS for the front end. The bridge implementation for the DOM engine supports eight environments, including C, Rust, Swift, Kotlin, C#, and WebAssembly.
- Cloud and Runtime Environment
- Built on Google Cloud Platform (around Cloud Run, Cloud KMS, Secret Manager and Vertex AI) and AWS, and developed and operated on Linux (Docker / Cloud Run) and Windows.
- LLMs and AI Agents
- We use the Claude, OpenAI and Gemini APIs according to purpose, adopting a configuration that does not depend on a single vendor. We have ourselves designed a development structure that uses AI agents to automate everything from requirements definition to implementation, verification and operation, and we are proving it in our own core business systems.
- Real-time audio signal processing
- float32 processing, PCM support up to 768kHz, and a low-latency implementation of 143 samples (3.24ms)
- AI agent design
- Building a group of agents that autonomously handle research, design, implementation and verification, and designing operations in which people retain judgment and final responsibility
- Encryption and licensing infrastructure
- Design of Ed25519 signatures (Cloud KMS), encryption of processing coefficients with ChaCha20-Poly1305, and a license revocation list
- Verification design
- A mechanism that guarantees quality mechanically, through conformance tests that check processing results bit by bit
- API/EDI Integration and Business Automation
- Core-system integration and automated pipelines using Amazon SP-API, Advertising API and others
- RAG, Evaluation, and Audit Infrastructure
- Building a mechanism that generates responses grounded in your own data and verifies their accuracy