Tip: select any text on this page to bring up the Ask tooltip and question the Synedrix OS assistant about it.
From human intent to governed execution.
A person states what they want done. Synedrix OS turns that into coordinated work across agents, workflows, systems, and the people who need to sign off.
Every new agent adds coordination work that never goes away.
Each project solves it again by hand, and the cost compounds with every agent added.
Every new agent still has to work with the agents and systems already running: who approved what, which policy applies, what happened before, and who is accountable if it goes wrong. Most teams solve this once per project, by hand, inside a single workflow. The coordination logicgets rebuilt every time, and it can't be reused by another team, carried into another domain, or explained to an auditor.
That unpaid coordination work is coordination debt, and it grows with every agent added. The fix is to treat coordination as a first-class layer of the platform, built once and shared by every project, so the debt stops compounding.
Two structures. One organisation.
The human org chart keeps accountability and judgment. The work graph coordinates execution inside the boundaries people set.
A traditional org chart does two jobs at once. It assigns authority, and it coordinates work. Synedrix OS separates them into two overlapping structures in one organisation.
The human org chart keeps accountability, strategy, and judgment: who owns outcomes, who sets policy, and who decides when stakes are high. The Synedrix OS work graph coordinates execution. Goals flow through it, get broken into work, and return as proposed actions or completed outcomes, inside the boundaries humans set.
The work graph holds no positional authority over anyone. It coordinates the work, and people keep the authority to decide.
- Board
- CEO
- Executive Leadership
- Business Owners
- Objectives
- Orchestration
- Agents + Humans + Systems
- Outcomes
One governed platform, however you choose to run it.
Self-hostable by design, with three deployment modes that let a regulated business choose where inference and data live.
Synedrix OS is a reusable multi-agent operating system. Specialised agents coordinate across workflows, tools, and data under centralised policies, approvals, and audit, so humans stay above the automation layer. Every component runs self-hosted in all three deployment modes (cloud-connected, self-hosted hybrid, air-gapped), with one recorded exception: the inbound event gateway is a managed service, called out on the platform page.
- Cloud-connectedInference on cloud APIs
- Self-hosted hybridCloud or local, per agent
- Air-gappedLocal inference
The architecture is complete. The platform is being built.
Synedrix OS is in active development. The architecture is complete and documented, and the reasoning behind it is published in the blog series. The platform is being built. The founding team is forming, and co-founders and early investors are welcome to build it with us.
Governed architecture
Policy, approvals, and audit are part of the orchestration layer, not added around it after the fact. Every consequential action is checked against policy before it runs, and every run leaves an audit trail.
Deployment sovereignty
Cloud-connected, self-hosted hybrid, or air-gapped: the same platform, with three choices for where your data and inference live and who controls them.
Trust as a structural property
Bounded authority, explainability, and human oversight are enforced by the platform itself, as a matter of how it's built.
One shared question per plane.
A way to read the platform by the job each part does. It is an organisational view; every agent keeps its technical family and layer.
Orchestration
Turns a stated objective into coordinated work across agents, workflows, and systems.
Control
Decides what may happen, who is authorised, and when a human must step in.
Intelligence
Surfaces what is happening now and what is likely next, with the evidence to act on it.
Memory
Holds what the organisation knows and governs what it learns, so context carries from one decision to the next.
Humans own outcomes and authority. Synedrix OS coordinates execution.
People keep the judgment. The platform carries the coordination.
Strategy, capital allocation, policy, and every high-impact commitment stay with people. What increasingly runs through the platform is the coordination around those decisions: breaking a goal into tasks, enforcing policy, and tracking execution.
See what changes and what doesn't →Twelve principles, each with an enforcement mechanism.
Each one names the field, agent, or validation rule that makes it true at runtime.
Governance that lives only in a document does not hold up under pressure. Every principle behind Synedrix OS is paired with the specific mechanism in the architecture that enforces it.
Read the principles →“Trust is the real product.”
Principle 10 (Trust is the real product): the ultimate output of the system is operational trust.
One platform. Four domains, so far.
Each pack adds domain-specific entities, workflows, and agents; the governed platform underneath stays shared.
A domain pack is a modular package that runs on Synedrix OS and adds what is specific to one business domain: its entities, workflows, and agents. Trading has a versioned specification (v2.0); Commerce, Ops, and Finance are scoped and in development. The governed platform underneath stays shared, so every pack inherits the same orchestration, policy, and observability instead of rebuilding them.
Explore the domain pack model →Twenty posts. Every architectural decision. Published weekly.
Start with the three posts below, then read the series in order.
Answers in plain language
Short definitions you can cite. Ask the assistant for deeper, source-grounded detail.
What is Synedrix OS?
Synedrix OS is the governed operating layer for agentic business systems: a reusable, self-hostable platform that coordinates specialised agents across workflows, tools, and data under centralised policy, approvals, and audit. It is a platform to configure, not a chatbot framework or a single-purpose automation tool rebuilt for every customer.
What problem does Synedrix OS solve?
Every new agent adds coordination work that never goes away. Most teams solve that once per project, by hand, inside a single workflow. Synedrix OS makes governance, orchestration, memory, tool mediation, and observability shared platform primitives instead of per-project reinvention.
What's the difference between Synedrix OS, a domain pack, and a business instance?
Synedrix OS is the reusable core. A domain pack configures it for one industry (entities, events, specialised agents, policies). A business instance configures it for one deployment (secrets, thresholds, approvers, tenant settings) without forking the platform underneath.
What are the core principles behind Synedrix OS?
Twelve principles govern how the platform stays trustworthy: among them, human accountability stays above the OS, domain logic stays out of the core, every workflow is observable end-to-end, agents have explicit authority boundaries, durable memory is committed only through governed paths, and every side-effecting step carries an idempotency key, with any step that cannot be made idempotent reconciled or escalated instead of replayed.
What deployment modes does Synedrix OS support?
Synedrix OS supports cloud-connected, self-hosted hybrid, and air-gapped deployments. The same core platform runs in each mode; inference location and external connectivity change, not the governance model.
How does Synedrix OS keep AI agents accountable and auditable?
Agents operate under explicit authority levels (recommend, approve, execute), policy-aware orchestration, escalation paths to humans, and full audit trails. No agent writes to production systems without orchestration visibility, and irreversible or regulated actions require declared approvals.
What are the canonical layers in the Synedrix OS architecture?
Twelve layers span human interface and goal/intent, through orchestration, agent runtime and agent registry, tool abstraction and memory, events, governance, observability, and finally domain pack and business instance configuration. Each layer has a clear owns / does-not-own boundary.
What are domain packs, and how do they extend the platform?
Domain packs add industry-specific entity models, event catalogs, specialised agents, tool adapters, policies, and workflow libraries on top of the shared OS. Trading is the first fully specified pack; Commerce, Ops, and Finance are registered scopes that inherit the same core primitives.
Follow the build
A weekly update on the build, with a sneak preview, and a notice at launch.
Follow the build