Vision · iryx · Capabilities · Stack · Philosophy · Direction · Contact
Neofenyx builds intelligent operational software for the AI era.
We work at the intersection of cloud engineering, AI, and spatial intelligence — applying them to the systems businesses actually run on: dispatch, fleets, routes, warehouses, customers, exceptions. Calm software that becomes more capable the longer it operates.
We are not an AI wrapper, an agency, or a consultancy. We are a product company building durable platforms for a generation of operations that will be increasingly assisted, observed, and reasoned about by machines.
A cloud-native operating system for same-day delivery and last-mile logistics. iryx unifies the operational surface of a modern courier business into one spatially-aware platform.
| Domain | What iryx does |
|---|---|
| Dispatch | Order intake, assignment, and live exception handling |
| Fleet | Vehicles, shifts, capacity, and driver orchestration |
| Routing | Plan, optimise, and re-route in response to real-world signal |
| Tracking | Live shipment state, ETAs, and customer-facing portals |
| Analytics | SLA, on-time, cost-to-serve, and operational intelligence |
| Automation | Workflow rules and AI-assisted operational decisions |
iryx is designed for operations that move. Geography, time, and exception are first-class concepts — not afterthoughts bolted onto a CRUD app.
We build, ship, and operate:
- AI-assisted operational systems — decisions, routing, exception handling
- GIS & spatial intelligence — maps, routing graphs, geofencing, telemetry
- Cloud-native platforms — distributed, observable, durable by design
- Operational analytics — turning movement and events into signal
- Enterprise SaaS — multi-tenant, identity-aware, audit-grade
- Workflow automation — from manual procedure to orchestrated system
- Data systems — modelled, governed, queryable
- Location-aware systems — geography as a first-class engineering type
A deliberately modern, deliberately boring stack — chosen for production lifespan over hype cycles.
| Layer | Tooling |
|---|---|
| Services & APIs | .NET, C# |
| Web clients | Angular, TypeScript |
| Cloud & infrastructure | Azure, Docker, distributed services |
| Data | SQL Server, PostgreSQL · PostGIS |
| Spatial & rendering | Mapbox, Leaflet, custom routing engines |
| Intelligence | Azure OpenAI, embeddings, retrieval, evals |
01 · Systems, not features. We design for the second year of operation, not the demo.
02 · Intelligence, applied. AI earns its place where it changes a decision, a route, or an outcome.
03 · Operational clarity. If you cannot see it, you cannot run it. Observability is a product surface.
04 · Boring where it matters. Durable tools, well-modelled data, calm interfaces — novelty where it earns its weight.
We are building toward operational software that observes itself, reasons about exceptions, and proposes the next correct action. Calmly. Continuously.
Three horizons:
- Platforms. Mature, multi-tenant operational systems with first-class spatial primitives.
- Agents. Domain-trained assistants that participate inside the workflow — not alongside it.
- Fabric. Operational data fabrics that let intelligence move freely between systems.
| Web | neofenyx.com |
| Product | iryx.io |
| Engineering | github.com/neofenyx |
| Inquiries | hello@neofenyx.com |