If you look at the most urgent challenges in enterprise AI today (autonomous agents, robotic process automation, and spatial intelligence) they all share the same bottleneck. We have language models that can talk brilliantly, but they struggle to interact safely with complex, physical systems.
At Intertangible, we saw this bottleneck coming years before “Agentic AI” became an industry buzzword. It is the reason our company exists.
Intertangible was formed through a deliberate, strategic merger of two highly specialized entities: Moto Japan, an automotive technology and Digital Transformation (DX) firm that operated from 2016 to 2019, and Domma, an AI research firm deeply specialized in linguistics and Natural Language Processing (NLP).
On paper, merging an automotive hardware DX consultancy with an NLP lab might have looked unorthodox in 2019. But structurally, it was an architectural necessity.
To understand why, we look to the foundational philosophy of Dr. Terry Winograd, Professor Emeritus of Computer Science at Stanford University. In the early 1970s, Winograd created SHRDLU, one of the first highly successful AI programs. His core breakthrough was proving that natural language processing is fundamentally hollow unless it is grounded in a “world model.” A computer cannot truly process a command like “move the red block” unless it understands the physical constraints of gravity, space, and the block itself. Winograd proved that language and physical reality cannot be engineered in isolation.
Our early case studies at Intertangible were built directly on this premise. When Moto Japan’s expertise in rigid automotive engineering collided with Domma’s fluid linguistic models, we began solving problems that standard software companies couldn’t touch:
- Predictive Maintenance through Semantic Telemetry: One of our earliest integrated projects involved cross-referencing messy, unstructured, natural-language driver reports with hard physical telemetry data from connected vehicles. By grounding the AI’s language processing in actual automotive physics, we built a system that could accurately predict mechanical failures before they happened, bridging human complaints with hardware diagnostics.
- Engineering DX for the Factory Floor: Automotive manufacturing is governed by massive, complex technical manuals and rigid safety protocols. We deployed early NLP models to ingest these thousands of pages of unstructured engineering documentation, transforming them into real-time, voice-accessible diagnostic workflows for technicians on the floor. We built a linguistic interface for a high-stakes physical environment.
These early case studies forged Intertangible’s core DNA.
Most AI companies today were born in the cloud. They understand software, text, and pixels. But because of our roots in automotive tech, we understand steel, supply chains, and physical consequences.
By merging Domma’s linguistic intelligence with Moto Japan’s physical systems engineering, we built a company uniquely equipped for the current era of AI. Today, as we deploy autonomous agents for global banks and rebuild supply chain architectures for pharmaceutical giants, we rely on the exact same foundation we established on day one: intelligent software must always be engineered to respect the unyielding realities of the physical world.