Modular by design
Explicit interfaces make upgrades possible. No one piece of hardware or software should define the architecture; that's for you to decide.
Engineering research + product development
Developing modular cybernetic systems on a traceable foundation.
Purpose
Savogelt is an early-stage engineering company working to unite biomechanics, software, sensory intelligence, and automated manufacturing.
SG is a way of building. We are developing a platform that defines interfaces, preserves evidence, separates observation from inference, and makes uncertainty visible. Traceability runs from the first requirement through fabrication, inspection, operation, and revision.
The goal is a set of interoperable components that can be maintained, revised, and used together.
Engineering philosophy
Explicit interfaces make upgrades possible. No one piece of hardware or software should define the architecture; that's for you to decide.
Every part should carry its own history, from requirements through manufacture and revision. That history gives the next engineer something to work from.
Real-world durability goes beyond wear. Understandable construction makes maintenance and repair possible throughout a part's working life.
Let the machine do its job and not yours. Human authority remains explicit as work is automated.
Interchangeable parts don't mean interchangeable states. Measurements, observations, derived data, inference, and commands each have their job and their place.
Standards earn their keep in the real world. Designs must be checked against operating conditions, tolerances, and maintenance needs.
Development programs
Savogelt's programs are designed to be useful on their own and to work together, from component design through manufacturing, inspection, and operation.
A modular mechanical and biomechanical platform centered on standardized interfaces, scalable geometry, serviceability, and traceable design rules.
SG-1 is in development as the first physical reference component, intended to test how the SG design language moves from specification into manufacture.
An operational-intelligence platform in development to connect physical systems, software, documents, sensors, and data into a live, evidence-backed model of an organization's environment.
Built from a working mobile application and OCR pipeline. The broader direction is to preserve sources and change history, make the environment searchable, and support controlled reasoning and automation.
An internal standards family and machine-readable registry for requirements, evidence, design, manufacturing, verification, change, software, risk, security, AI, metrology, and control.
SG is a living language for human engineers and automated systems. Its rules and records are designed to evolve with the work.
A local, tool-first AI environment for engineering and fabrication: models coordinate work, deterministic tools perform calculations and measurements, independent checks verify results, and operational systems keep their own state.
This includes orchestration across CNC, additive manufacturing, inspection, technical-document ingestion, and future automated work cells. These workflows are intended uses of Savogelt's Scryer deployment.
Scryer's generic model of an organization and its state: assets, machines, software, networks, documents, workflows, evidence, and their relationships. Savogelt operates its own reference deployment.
SEM is being designed to distinguish what should be, what is, and what is changing. Savogelt's standards and internal governance belong to its deployment.
Method
The engineering record is part of the product. Each transition should preserve enough context to explain what changed, why it changed, who or what authorized it, and what evidence supports it.
Define the problem, intended use, boundaries, assumptions, and exclusions.
Collect observations, measurements, references, constraints, and operating conditions.
Translate evidence into testable functional, interface, safety, manufacturing, and quality requirements.
Build explicit architectures, parameters, interfaces, calculations, models, and decision records.
Record what was actually made, with materials, machines, programs, tools, process states, and deviations.
Test the physical and digital result against released requirements using controlled methods and evidence.
Observe real performance, maintenance, failure, repair, configuration changes, and edge cases.
Feed validated lessons back into requirements, models, design rules, processes, and future revisions.
Company state
Savogelt is building its engineering foundation and first demonstrators. The status below separates working prototypes, development work, and planned systems.
Collaboration
Savogelt is interested in manufacturing partners, research collaborators, component suppliers, standards work, machine-vision tooling, robotics, metrology, and practical fabrication knowledge.
Contact
For engineering, research, manufacturing, supply, or collaboration inquiries, email Savogelt directly.