Skip to content
    Back to Agentic IT Operations
    Agentic IT OperationsStartupOSS Enterprise Search

    Onyx (ex-Danswer)

    Open-source AI search engine for internal IT docs and knowledge

    Mkt Cap / ValOpen Source
    RevenueEarly Stage
    Growth+150% YoY
    Mar 2025: $10M seed (Khosla, First Round); rebranded to Onyx
    Open-source AI search and assistant that unifies internal knowledge with self-hosting and LLM choice.
    Analyst take · Competitive edge

    SWOT Analysis

    Strengths
    • Open-source: self-host, choose any LLM, full data residency control
    • Connects Slack, Drive, Confluence, GitHub, Salesforce and more
    • Granular access controls suit security-conscious buyers
    • Builds custom agents and runs deep research over company data
    • Strong references including Netflix and Thales Group
    Opportunities
    • Win enterprises wary of closed SaaS search on sensitive data
    • Ride agentic-RAG and internal-knowledge-assistant demand
    • Expand connector catalog and managed-cloud offering
    • Leverage open-source community for adoption and contributions
    Weaknesses
    • Self-hosting adds deployment and maintenance burden
    • Open-core monetization must convert free users to paid
    • Small seed-stage team versus large enterprise-search incumbents
    • Search quality depends on connector coverage and data hygiene
    Threats
    • Glean, Microsoft Copilot and Google bundling enterprise search
    • Foundation-model vendors adding native enterprise connectors
    • Open-source forks pressuring commercial differentiation
    • Buyer fatigue across crowded internal-search market

    User Sentiment

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

    What users love
    • Open-source flexibility with self-hosting and LLM choice
    • Broad connector set unifying scattered knowledge
    • Granular permissions that respect source access controls
    • Custom agents and deep research over internal data
    Common complaints
    • Self-hosted setup and upkeep require engineering effort
    • Answer quality varies with connector and data quality
    • Smaller vendor support footprint than incumbents

    Customer Profile

    Who buys this

    Typical segments

    EnterpriseSecurity-conscious orgsEngineering teams

    Typical buyer

    Head of IT / Platform or knowledge-management lead

    Top use cases
    1. 1Natural-language search over internal docs
    2. 2Custom internal AI assistants and agents
    3. 3Deep research across company knowledge

    Future Focus Areas

    1

    Deeper agentic workflows and actions

    2

    Managed-cloud and enterprise tiers

    3

    Expanded connector ecosystem

    4

    Stronger permission-aware retrieval