Features

Everything in the live stack.

Ingest, transcode, package, deliver, record, encrypt, caption, chat — every layer of the Kardolive media plane is ours. No vendor in the middle. No surprise breakage.

Ingest

Every protocol your hosts already use.

RTMP, RTMPS, SRT, WHIP, WebRTC, HLS pull, RTSP — all routed through one ingest layer with shared auth, monitoring, and quota. Mobile SDKs use the same path your browser hosts do.

WHIP (WebRTC)
Browser-native, sub-second, no client install. Default for low-latency creators.
RTMP / RTMPS
Broadcast software fans — works with what your operators already know.
SRT
Long-haul ingest tolerant of bad networks. The choice for outdoor and remote.
HLS pull / RTSP
Pull from existing CDN feeds or RTSP cameras. Useful for surveillance and IoT.

Real-time SFU

Massive rooms without the lag.

Sub-300ms glass-to-glass with simulcast + SVC. Per-tenant SFU pods, sticky-routed by geo and load, so a noisy neighbor never starves you.

Simulcast + SVC
Publishers send multiple layers; receivers pick what fits their network.
Geo-aware routing
Placement service picks the closest SFU based on participant geo + load.
Per-tenant pods
Heavy meetings on one tenant don't steal CPU from another.
Up to 50K viewers / cluster
Horizontal scale — add a cluster, get another 50K seats.

Transcoding

Every ladder rung, every codec.

CPU + GPU transcoders deliver 1080p60, 1080p, 720p, 480p, 360p, 240p, audio-only. H.264, H.265, AV1, VP9. Opus + AAC audio. Spot-priced for VOD, on-demand for live.

Adaptive bitrate
Auto-ladder per stream. Or override per-tier for content with specific needs.
AV1 + HEVC
Higher quality at half the bitrate. AV1 the default once devices catch up.
GPU encode
NVENC workers for live; CPU spot for VOD batch transcoding.

Packaging + playback

HLS, LL-HLS, DASH, WHEP, CMAF.

Pick the playback protocol that fits the platform. Per-segment encryption, signed playback URLs, multi-CDN ready.

LL-HLS for the long tail
1–2 second tail latency at replay-grade quality.
WHEP for the front row
Sub-second WebRTC playback for interactive viewers.
Signed playback
Per-tenant CloudFront key groups; scope URLs to user + window.
Multi-CDN abstraction
Add Akamai/Fastly later if a customer mandates it. Same control plane.

Recording + VOD

Every live becomes a shoppable replay.

Cloud recording composites SFU streams to a single MP4, then transcodes a full VOD ladder. Replays carry timestamped product pins for shoppable scrubbing.

Layout-aware compositing
Grid, speaker, picture-in-picture, custom canvas layouts.
Chapter markers
Auto-generated from product pins; viewers scrub by moment.
Retention tiers
Hot / IA / cold storage per policy. Auto-expire by tenant rule.

DRM

Premium content protection out of the box.

Widevine, FairPlay, PlayReady — our own license servers, your keys, your control. No third-party DRM SaaS in the path.

Widevine
Android + Chrome desktop.
FairPlay
Safari, iOS, tvOS.
PlayReady
Smart TVs, Xbox, Edge.

Chat + signaling

Real-time chat at every layer.

Built-in chat, presence, polls, Q&A, reactions, gifts, super-chats. Threaded for meetings, fanout for livestreams, persistent for spaces.

Sharded by room
Horizontal scale to millions of concurrent chat participants.
Moderation
AI + human moderation, blocklists, profanity filter, rate limit.
Reactions + gifts
Per-tenant gift catalog. Monetized with one switch.

Live transcription + translation

Real-time captions in 29 languages.

Captions in the stream. Translations in the chat. Both run on our own infra — no caption-as-a-service vendor in the loop.

Burnt-in or sidecar
Choose between hardcoded captions or WebVTT sidecar tracks.
Per-viewer language
Each viewer picks their own. Auto-detect from browser.
RTL support
Arabic, Urdu, Persian — typography, ligatures, layout all correct.

Stop renting your media plane.

Start free Talk to sales