Understand the request
Use natural language, current system state, recent events, and relevant context to work out what the user is asking for.
Independent software startup
ORION connects AI decisions to the systems that have to carry them out.
TylerInnovated is building ORION, an AI orchestration platform that turns a request into a controlled workflow, watches the systems involved, and reports whether the intended result actually happened.
- Current status
- Working private deployment
- Public evidence
- Interactive demo using simulated data
- Core technology
- AI reasoning, tools, and verification
Why ORION exists
AI can understand a request. The harder part is carrying it out safely and knowing whether it worked.
A normal assistant may send a command and report success. ORION checks live state, chooses a controlled tool or workflow, and looks again after the action. If a service is offline or the result cannot be confirmed, the system should say so plainly.
Route the action
Send the task through approved automations, APIs, local services, and connected devices instead of giving the model direct control.
Check the result
Compare the requested outcome with observable system state and report success, failure, partial completion, or uncertainty.
Product evidence
ORION is working software, not a concept-only landing page.
The private development deployment includes a Node.js service, multiple operational interfaces, system diagnostics, quick actions, event history, and status views for Home Assistant, n8n, devices, cameras, and local services. The public demo mirrors those workflows with sample data and no access to private infrastructure.
Operational proof
A working application stack with explicit verification rules.
The current ORION V1 codebase separates the browser interface, the action router, live system state, evidence collection, and outcome verification. It also includes deterministic tests and release checks for the status contract, action routing, security verification, notification policy, and public-link handling.
Running backend
A local Node.js service owns system truth and exposes stable status, events, links, and action contracts.
Real integrations
Home Assistant provides connected-state data while n8n handles defined automation workflows.
Checked outcomes
Actions are not reported as successful until the requested postcondition can be observed.
Public demonstration
A safe browser-only environment lets reviewers explore the product without reaching private systems.
Initial focus
The first ORION deployment is a connected home and homelab.
That environment is a useful proving ground because it contains real devices, changing system state, several vendors, local services, automations, and actions that should not be treated as successful without a check.
What it tests today
- Service outages and unavailable integrations
- Device state that changes outside ORION
- Automations that complete only part of a task
- Security checks that require clear confidence and limits
Where the product can grow
- Additional homes and personal infrastructure
- Small offices with connected tools and devices
- Operational environments that need one control layer
- Remote services supported by secure cloud infrastructure
How it works
A request moves through six clear steps.
ORION keeps AI reasoning separate from the systems that perform the action. The model can interpret and plan, but approved tools handle execution and ORION checks the result.
- RequestNatural language, a dashboard action, a routine, or a system event
- ContextCurrent state, recent events, rules, preferences, and relevant memory
- AI decisionIntent, planning, explanation, and selection of a structured tool
- Controlled actionApproved workflows, APIs, local services, and automations
- System responseHome Assistant, n8n, devices, cameras, and connected services
- Check and reportOutcome checks, exceptions, confidence, and a clear result
Development status
What is working now, and what is being built next.
Current product surfaces
- Browser based dashboard, Command Center, and Guard Mode
- System health, diagnostics, alerts, and attention views
- Quick actions and control surfaces for common routines
- Status visibility for Home Assistant, n8n, devices, and services
- Clear handling of offline, unavailable, and unconfirmed state
Intelligence and deployment work
- Persistent memory and semantic retrieval
- Deeper model planning and structured tool selection
- More policy checks and outcome verification rules
- Multimodal event and environmental understanding
- Secure remote services and cloud supported features
- Repeatable deployment in additional environments
Founder and public identity
TylerInnovated is founded and led by Tyler Hanny.
Tyler leads product development, system architecture, interface design, and the current ORION implementation. Public third-party profiles are linked below for independent verification.
Tyler HannyFounder and product developer
About TylerInnovated
An independent AI startup turning ORION into a repeatable product.
TylerInnovated is led by Tyler Hanny in the United States. The current work is focused on product development, technical validation, and turning ORION into a system that can be deployed beyond its first environment.