Video Wall Voice Control SDK
Multi-screen array for public spaces, control rooms, or events
How BotSurf works on a Video Wall
BotSurf runs on a Video Wall as an on-device AI voice agent, not a fixed list of matched commands. Voice control of displayed content, switching views, searching information. Because the underlying engine is a real AI model rather than a command-matching layer, it handles phrasing variation, multi-step requests, and follow-up questions the way a person actually talks — not the rigid "say this exact phrase" pattern older voice integrations require.
The manufacturer opportunity
Public venue and control room voice interaction. This is a genuine AI differentiator, not a rebadged wake-word matcher — Video Wall manufacturers can market real on-device AI reasoning, multi-step task handling, and privacy-preserving local processing, positioning against competitors still shipping fixed-command "voice control" as if it were the same thing.
Hardware & memory baseline
The floor TVs & Displays devices need to clear before integration; exact figures are confirmed per project once we know your chipset and target markets.
Minimum RAM
2 GB, 4 GB recommended for simultaneous app switching, playback and AI agent execution
Minimum storage
8 GB available for the OS-level install, OTA headroom, and the preloaded content catalog
CPU architecture
ARM Cortex-A55 or better; the AI wake-word and intent models run comfortably on TV-class SoCs already shipping today
Microphone array
Far-field 2+ mic array for always-listening use; single mic (remote-embedded) supported for push-to-talk
Network
Wi-Fi or Ethernet for setup, streaming app control and OTA updates; on-device intent recognition works with an intermittent connection
Display output
Required — visual search results, agent responses and app switching are shown on-screen, not just spoken
Chipset compatibility
Pre-optimized for Amlogic and MediaTek (the dominant Smart TV/set-top SoC vendors), plus Qualcomm Snapdragon for premium tiers