Skip to content
Signal Capital
Applied intelligence
All domains
04Embedded software edge

Applied intelligence

Intelligence is most valuable when embedded in real-world systems. We invest in computation that operates in physical and biological domains — not generic AI, but purpose-built systems with measurable impact.

Thesis

The most defensible AI and software companies are not building general-purpose tools. They are embedding computation deeply into specific industrial, energy, and biological workflows where domain expertise, proprietary data, and real-world feedback loops create compounding advantages. We look for applied intelligence companies where the software cannot be separated from the domain.

Where we focus
01

Industrial AI

Machine learning and optimization systems for manufacturing, process control, predictive maintenance, and quality assurance in physical environments.

02

Scientific computing

Computational platforms for drug discovery, materials science, and climate modelling that accelerate research timelines.

03

Edge intelligence

On-device and near-device computation for real-time decision-making in industrial, energy, and infrastructure systems.

04

Systems software

Infrastructure software, simulation tools, and digital twins for complex physical systems across energy, manufacturing, and logistics.

Why it matters
  • Vertical workflows often capture more value than horizontal tools when data and feedback loops are proprietary.
  • Latency, safety, and uptime push inference and control toward the edge in many industrial settings.
  • European buyers increasingly expect transparency, auditability, and fit with existing OT/IT stacks.
  • Sustainable advantage comes from deployment depth — not model scale alone.
How we frame it
EmbeddedSoftware inseparable from domain workflows and data
MeasurableKPIs tied to throughput, yield, uptime, or risk
Edge-readyArchitectures suited to real-world constraints
Next domain

Research platforms

Infrastructure and tools that accelerate discovery, validation, and deployment cycles.