Describe what you want in plain engineering terms, or compose it visually on a canvas. RAN Flow turns your RAN logic into a running, auditable rApp. In days, not months.
Building an rApp today needs radio knowledge and software skills at once, and neither expert has the right tool for the other's half.
“I know exactly what policy to push, but I can't write the code to do it.”
Strengths
Friction
RAN Flow
bridges the gap
“I can build the pipeline, but I don't know what values to set.”
Strengths
Friction
Concept to deployable rApp.
You control rApp delivery.
Every rApp is an auditable graph.
SMO-portable rApp packaging.
Days, not months, to your first rApp
Concept to deployed rApp on your own SMO, with no software team in the loop.
Two ways in, one result. Use whichever fits how you think, and RAN Flow meets you there.
Write the outcome in plain engineering language and RAN Flow will draft the workflow for you to review. No syntax, no SDK.
# intent
"Shed load from congested cells
when PRB stays above 80% for 5 min."
# RAN Flow drafts
Drag triggers, data readers, logic gates and actions onto a canvas. 22 node types across four categories, so you see the whole rApp at a glance.
# visual canvas
Either way, you get the same thing: a graph you can read, dry-run, and sign off on before it touches the network.
Bring a use case: a KPI target, an optimization idea, a legacy SON feature to modernize. Our Forward Deployed Engineers embed with your team, build it on RAN Flow, validate it against your data, and deploy it on your SMO. You keep it: a readable graph your engineers can edit, with no lock-in.
A KPI goal or optimization idea in plain engineering terms. No spec, no code.
We compose the rApp on RAN Flow, validate against your data, and walk it through your change process.
It ships on your SMO as an editable graph, and becomes a template your team reuses on its own.
Built on the same platform you'd use yourself, so every engagement leaves you more self-sufficient, not less.
On a live national network, nobody approves code they can't read. RAN Flow keeps every rApp inspectable, testable, and under human control.
Every rApp is a graph you can open and audit, never generated code you have to trust blind.
Validate logic against mock RAN data and a 12-rule checker. Catch misconfigurations before they ship.
You review the workflow, change it, and decide when it goes live. Nothing reaches the RAN on its own.
A1 policy actions ship with built-in bounds. PRB targets stay inside 5–95%, so an rApp can never black out a cell.
Drag-and-drop workflow builder with 22 node types across 4 categories. See your rApp logic at a glance.
Turn a plain-language goal into a draft workflow, ready to review and refine.
Train forecasting models on your own KPI history and version them in the platform, so every rApp calls the same one.
Push SLA and load-balancing policies to the Near-RT RIC via the A1 interface. Guard rails built in.
Forecast a KPI 15 minutes to 6 hours ahead with LSTM, ARIMA, or your own ONNX model, then branch on what the forecast says before the degradation lands.
Connect to live metrics like throughput, PRB usage and latency, then trigger actions when thresholds are crossed.
If/else branching with true and false output paths. Build complex decision trees without code.
Deploy workflows as persistent background rApps. Full lifecycle management: start, pause, stop.
Test workflows against mock RAN data before touching production. A 12-rule validator catches misconfigurations early.
Every workflow compiles to a standalone Python script you can read, diff, and run outside the platform.
Your rApps aren't tied to one vendor's stack or one operator's platform. Every RAN call goes through a connector layer, so the same workflow moves between SMOs. When you add or switch one, your logic comes with you.
| FEATURE | THE PLATFORMRAN Flow | LEGACY APPROACH“Write the rApp yourself.” |
|---|---|---|
| Time to first rApp | Days, not months | Weeks to months |
| Required skills | RAN domain knowledge only | Python, K8s, SDK |
| Validation | Automated, 12-rule validator | Manual, trial and error |
| Auditability | Readable graph, always | Opaque generated code |
| Multi-vendor | EIAP and O-RAN SC, portable by design | Vendor-locked |
| Maintenance | Update once, deploy anywhere | Re-code per platform |
| Dry run and testing | Full dry-run before deploy |
CAPABILITIES AS SHIPPED TODAY
In private beta with a small group of operators. Leave your work email and we'll be in touch.