FAQ

How Rilla works

How does Rilla work in simple terms?

Rilla allows your live streaming viewers to safely relay encrypted video segments using their upstream bandwidth. That improves video stability and reduces load on your CDN when peer-assisted delivery is suitable.

A lightweight Performance P2P SDK integrates into your streaming stack at the media-segment layer. An AI Orchestrator intelligently decides when and how viewers share data in real time, and always falls back to your CDN when needed.

Rilla augments your existing CDN; it does not replace it. Evaluations measure CDN deflection and playback guardrails on representative live traffic before broader rollout.

Learn more: What is Rilla, How It Works, CDN Deflection.

Does Rilla replace my CDN?

No. Rilla augments your CDN; it does not replace it.

Your CDN remains the primary delivery mechanism and instant fallback path throughout playback. Rilla adds a peer-assisted layer so eligible viewers can relay encrypted video segments to others watching the same live content. Traffic stays on your CDN when peer delivery is unsuitable, when no eligible peers are available, or when your rules require CDN-only delivery.

A POC is not a CDN migration. It validates deflection, playback safety, and integration fit while your existing CDN, origin, encoding, DRM, and player workflows stay in place.

Learn more: What is Rilla, CDN Deflection, POC / Trial Guide.

How does Rilla fit into my existing streaming stack?

Rilla integrates at the player and segment-request layer, not at the origin or encoding pipeline.

The Performance P2P SDK sits between your player, your CDN, and eligible peers. Your player keeps requesting video from your CDN; the SDK intercepts eligible segment requests and applies Rilla delivery decisions. When peer delivery is not suitable, playback stays on your CDN.

You do not need to replace your CDN, multi-CDN setup, DRM system, entitlement layer, or ad workflow. Rilla works alongside them. Authentication, entitlements, and access rules remain authoritative.

Learn more: How It Works, Performance P2P SDK, Broadcast Grade.

How is Rilla’s use of P2P different from historical P2P?

Traditional P2P models focused on static file transfer. Rilla is engineered for the dynamic and unpredictable demands of real-time video streaming.

Key differences:

Real-time orchestration: Rather than basic trackers or swarming, Rilla uses real-time orchestration based on dynamic network conditions. That handles significant network volatility without disruption to the viewer experience.
Stable high-throughput quality: All peers are not treated equally. If a peer loses the ability to contribute, that node immediately and safely falls back to your CDN.
Integrated for streaming: Rilla acts as a fully integrated layer that handles adaptive bitrates (ABR), content protection (DRM), authentication, ads, and rollout controls that previous P2P solutions largely ignored in production workflows.

Learn more: Broadcast Grade, Real-Time AI Orchestration, P2P Network.

How does Rilla use AI?

Rilla uses an AI Orchestrator service that observes network conditions to optimize video delivery between peers. It does not deliver video data; the Peer SDK executes coordination decisions in the player.

Rilla uses deep reinforcement learning (DRL) to train the system in a purpose-built, patent-pending simulation environment. During training, the DRL agent learns an optimal, real-time policy for complex scenarios such as peer selection, load distribution, and CDN offload ratios.

During live streaming, those learned models run as low-latency inference at a fraction of the training cost, enabling Rilla to orchestrate the network with predictive accuracy and high reliability.

Learn more: Real-Time AI Orchestration, AI Orchestrator.

Cost and CDN deflection

How does Rilla reduce the cost of delivery?

Rilla reduces content distribution costs by safely deflecting traffic away from your CDN using viewers’ available upstream bandwidth.

Expected deflection and savings are measured during representative live traffic. They depend on audience overlap, platform eligibility, playback guardrails, traffic shape, CDN contracts, and rollout scope. For target ranges and what drives results, see What CDN deflection rates can we expect? below and CDN Deflection.

When deflection is commercially meaningful, fewer CDN-served bytes can translate into lower delivery cost per event or per viewer. A POC validates measured deflection and QoE against your CDN-only baseline before broader rollout.

Learn more: Cut Delivery Cost, CDN Deflection, POC / Trial Guide.

What CDN deflection rates can we expect?

CDN deflection is the percentage or volume of played video bytes delivered through Rilla’s peer-assisted layer rather than directly from your CDN. It is how delivery efficiency ties to commercial outcomes.

In suitable live-streaming conditions, 60–90% CDN deflection is a common POC target range. The measured result depends on peer density, concurrency, platform eligibility, network conditions, rollout scope, and playback safety rules.

Typically, in a network where 30% of the audience can consistently contribute bandwidth during a live event, Rilla can achieve 80–90% CDN deflection. Your measured result will still depend on the factors above.

Deflection is not pursued at any cost. Rilla keeps CDN fallback available and measures startup time, rebuffering, and fallback behavior against CDN-only controls. Audience overlap, ad personalization, DRM paths, and ABR behavior can all affect how much traffic can safely move through peers.

Learn more: CDN Deflection, Cut Delivery Cost, Quality Under Load.

Broadcast-grade delivery

Does Rilla affect playback quality (QoE)?

Yes. Rilla is designed to protect and enhance playback quality at scale, especially when a large concurrent audience increases usable peer capacity on the same live stream.

The AI Orchestrator is trained to maintain integrity and stability and to reduce rebuffering under load. The larger the concurrent audience on the peer-assisted network, the more stable delivery can become: more viewers connected to reliable sources of traffic can improve the overall experience through a network effect.

Peer-assisted delivery is still conditional. The Orchestrator and Performance P2P SDK evaluate peer suitability, buffer health, and playback deadlines before using peer delivery. Deflection only counts when startup time, rebuffering, playback errors, and fallback frequency stay within agreed guardrails, measured against CDN-only control groups during a POC.

Learn more: Quality Under Load, CDN Fallback, Observability and Monitoring.

Will viewers notice that Rilla is enabled?

Viewers should not notice a product change. They keep the same player and app; Rilla works behind the segment-request layer.

Playback stays on a CDN-capable path. Peer-assisted delivery is used only when eligibility, timing, and policy allow, with CDN fallback when peer delivery cannot meet playback deadlines.

For how QoE and guardrails are measured during a POC, see Does Rilla affect playback quality (QoE)? above and Quality Under Load.

Learn more: How It Works, CDN Fallback.

How do you keep delivery stable when viewers constantly join or leave?

With continuous monitoring and real-time orchestration, every viewer’s upstream capacity and network conditions are managed in real time.

P2P networks can be unstable because they rely on the upstream bandwidth of each contributing node. If local network conditions or contention become an issue, Rilla accommodates that in the topology. If viewers leave abruptly, Rilla’s orchestration instantly adjusts the network within buffer-aware deadlines so quality of experience is maintained, including fallback to your CDN when peer delivery is no longer suitable.

Learn more: AI Orchestrator, Capacity on demand, Broadcast Grade.

Does Rilla work with DRM and content protection?

Yes. Rilla supports encrypted segment relay without replacing your existing content protection model.

DRM license requests, authentication, entitlement, and playlist-level access control bypass peer delivery. If a viewer is not allowed to access a playlist, they cannot participate in the related peer session. Manifests are not distributed between peers.

Integration scopes segment eligibility and fallback so peer delivery applies only where authorization and delivery semantics allow.

Learn more: Broadcast Grade, Security & Privacy, CDN Fallback.

How does Rilla work with ad insertion technology?

Rilla separates shared live video segments from personalized or CDN-only requests.

SSAI, CSAI, and other ad workflows can keep per-user ad or variant traffic on your CDN. Shared segments can still use peer-assisted delivery where media overlap is predictable. Personalized ad breaks, ad markers, DRM license requests, and non-whitelisted traffic can bypass Rilla by design.

Heavy ad personalization can reduce segment overlap and deflection. Start with content windows where media overlap is easiest to reason about, then validate personalized breaks in the POC.

Learn more: Broadcast Grade, CDN Deflection, POC / Trial Guide.

Platforms and scale

Which platforms and players are supported?

Rilla uses one Peer SDK model per platform, with player-specific integration guides on top.

Available today:

Web / Browser: @rillanetwork/peer-sdk-web with guides such as Bitmovin React and ExoPlayer (Android TV and mobile) .

Coming soon (per current docs):

Android SDK (platform-level)
Apple SDK (iOS / tvOS)

POCs typically start with one representative platform and player, then expand after validation. Confirm your target environment with Rilla before scoping the trial.

Learn more: Performance P2P SDK, Player Integrations, POC Checklist.

How does Rilla scale to millions of viewers?

Rilla’s architecture is engineered for massive, dynamic audiences through horizontally scalable orchestration.

Massive horizontal scaling: Rilla’s backend is designed to scale horizontally up to multiple millions of viewers within regional boundaries. The AI orchestrator is implemented on a proprietary stack designed for coordination-free operation at scale.

Network effect: Every Rilla-enabled viewer strengthens the network, automatically joining a regional relay layer. This expands delivery capacity as the audience grows, allowing streams to scale beyond the limitations of traditional CDN-only planning for peak windows.

Low-latency AI: Rilla’s AI model is optimized for low-latency inference, making peer topology decisions in tens of microseconds. It supports over 10,000 matchmaking decisions per second per shard.

Dynamic audience scaling: The lightweight signaling service accommodates tens of thousands of new viewer connections per second, handling the most abrupt surges in audience size.

Learn more: AI Orchestrator, Architecture, On-Demand Capacity.

POC trials

How can I test Rilla safely and prove my business case?

A Rilla POC validates whether peer-assisted delivery can deliver meaningful CDN deflection at scale while playback and system metrics stay within agreed thresholds , and whether integration fits your player, CDN, DRM, ads, and rollout workflows before production scaling.

In scope: CDN deflection and offload under real concurrency, playback safety, QoE guardrails, integration complexity, and a clear measurement framework.

Out of scope: Replacing CDN infrastructure, redesigning the delivery architecture, full device coverage, or treating the POC as full production rollout.

How to run it:

Define goals, target platform, player, region, and concurrency (see POC Checklist).
Integrate the SDK and confirm telemetry, CDN headers, and feature flags (Testing Approach).
Run A/B tests (CDN-only vs CDN + Rilla) on a representative live stream.
Measure deflection, peer-delivered and CDN-delivered volume, fallback frequency, and QoE deltas vs control.
Decide on phased rollout only when results meet agreed thresholds.

Rilla is in co-development with major broadcasters; POCs define a joint delivery path with a shared feedback cadence before broader deployment.

Learn more: POC / Trial Guide, Example POC Roadmap, Testing Approach.