Skip to content
    Back to Agentic IT Operations
    Agentic IT OperationsStartupLLM Agent Platform

    Ultravox (ex-Fixie)

    Platform for building and deploying LLM-powered IT automation agents

    Mkt Cap / ValPrivate
    RevenueEarly Stage
    Rebranded to Ultravox; $17M for open-weight voice SLM
    Speech-native open-weight SLM skips speech-to-text for ~150ms latency, real-time voice agents
    Analyst take · Competitive edge

    SWOT Analysis

    Strengths
    • Direct audio-to-LLM design removes ASR stage for low latency
    • Open-weight SLM on Hugging Face builds developer trust and adoption
    • Managed Realtime platform offers build-and-deploy path for agents
    • Speech-native handling enables more natural, fluid conversation
    • Backed by Redpoint, Madrona, Zetta and SignalFire
    Opportunities
    • Surging demand for real-time voice agents in support and ops
    • Win developers wanting open, self-hostable voice infrastructure
    • Embed as the voice layer inside broader agent platforms
    • Differentiate on latency as voice UX expectations rise
    Weaknesses
    • Modest $17M funding versus well-capitalized voice-AI rivals
    • Narrow voice focus limits cross-sell beyond conversational use
    • Smaller team and ecosystem than incumbent CX platforms
    • Open-weight model can aid competitors as well as customers
    Threats
    • OpenAI, Google and ElevenLabs pushing realtime voice APIs
    • Foundation-model vendors bundling native speech capabilities
    • Commoditizing voice stack compressing pricing
    • Larger players matching latency with deeper resources

    User Sentiment

    Synthesized from G2, Gartner Peer Insights, and analyst review data.

    What users love
    • Genuinely real-time, low-latency conversational responsiveness
    • Open weights allow self-hosting and customization
    • Bypassing transcription yields more natural dialogue
    • Developer-friendly model plus managed platform options
    Common complaints
    • Smaller ecosystem and tooling than incumbent voice platforms
    • Self-hosting open weights demands real ML infrastructure effort
    • Enterprise features still maturing relative to scale

    Customer Profile

    Who buys this

    Typical segments

    DevelopersStartupsVoice-AI builders

    Typical buyer

    Engineering lead or founder building voice products

    Top use cases
    1. 1Real-time voice agents and assistants
    2. 2Speech-native conversational interfaces
    3. 3Self-hosted low-latency voice apps

    Future Focus Areas

    1

    Multilingual speech support

    2

    Enterprise telephony and contact-center integrations

    3

    Improved model size and accuracy

    4

    Tooling for agent orchestration around voice