Cedar

What do they actually do

Cedar builds an open‑source React framework (Cedar‑OS) for adding an in‑app AI copilot to web applications. Developers install the cedar‑os package, use prebuilt UI components, connect to an LLM provider, and hook the copilot into frontend state and actions to guide users through workflows inside the product, not just in a chat box (GitHub, NPM, Getting Started).

Out of the box, it supports streaming chat, tool calls, state access, and “agentic actions” that can trigger product behaviors and render custom UI, so the copilot can help users complete tasks step‑by‑step inside the app (Architecture, Chat overview, Agentic actions).

Cedar also includes patterns for human‑in‑the‑loop review/approval and positions its product to help teams understand user intent and product‑market fit signals from in‑product interactions (Human‑in‑the‑loop, YC company page).

Who are their target customer(s)

• Product teams at SaaS companies with powerful but confusing UIs: New users don’t reach value quickly because they don’t know which features matter or how to use them; teams need a way to guide each user from signup to a first success without manual hand‑holding.
• Engineering teams building React web apps that want an in‑app copilot: Connecting an AI agent to frontend state and actions, streaming responses, and making the interaction reliable in production is time‑consuming to build and maintain.
• Customer success and support teams trying to reduce repetitive tickets: They repeatedly answer “how do I do X?” questions and escalate to product/engineering; they need a system that can answer or route with human oversight when necessary.
• Growth and early GTM teams seeking user intent signals and PMF analytics: They lack scalable visibility into what users are actually trying to do in the product, making it hard to personalize onboarding or re‑engage churned segments based on intent.
• UX/design teams for complex productivity or developer tools: Tooltips, docs, and generic chatbots don’t teach real workflows in context; they need contextual guidance that shortens learning curves and helps users complete multi‑step tasks inside the app.

How would they acquire their first 10, 50, and 100 customers

• First 10: Leverage founder and YC networks to run paid, hands‑on pilots with 10 SaaS products that have complex UIs; embed Cedar to guide one key workflow and measure activation lift, with weekly check‑ins and a case study per pilot.
• First 50: Publish ready‑to‑run React templates (onboarding, billing, search), run stipend‑backed installs via GitHub/Slack/Discord, and hold office hours; use early case studies for targeted outreach and convert via low‑friction paid pilots or short contracts.
• First 100: Build a self‑serve path (CLI/UI install, templates, playbooks) plus a paid support/tuning tier; add agency/platform partners who can recommend and install Cedar, and use how‑to content and measured case studies to convert trials to paid.

What is the rough total addressable market

Top-down context:

Cedar sits at the intersection of in‑app AI copilots, product onboarding/guidance, and support/CS deflection for B2B SaaS and developer tools. The broader markets for onboarding/product guidance and support chatbots are each multi‑billion dollar categories, with in‑app AI expanding their scope.

Bottom-up calculation:

If Cedar sells a commercial support/analytics tier averaging $10k–$25k ACV to 10,000 mid‑market+ SaaS apps that have complex UIs and active onboarding/support needs, TAM would be roughly $100M–$250M. Expanding to 20,000 such apps lifts TAM to ~$200M–$500M.

Assumptions:

• Initial focus is B2B React web apps with complex workflows (mid‑market and up).
• Pricing is primarily per‑account ACV for support/analytics and enterprise features, while the core remains open‑source.
• Penetration estimates exclude long‑tail small apps and non‑React stacks in the near term.

Who are some of their notable competitors

• CopilotKit: Open‑source React UI and infrastructure for in‑app AI copilots and agents; similar developer audience building embedded assistants inside apps (GitHub).
• assistant-ui: Open‑source TypeScript/React library for conversational AI interfaces; popular for quickly adding customizable chat UIs in apps (GitHub).
• Appcues: No‑code product onboarding and in‑app guidance; widely used to drive activation with tours and checklists (a non‑AI baseline many teams start from).
• Intercom (Fin): Customer service platform with AI assistant (Fin) for in‑app and web chat support; addresses the support deflection use case at scale within apps.
• Pendo: Product analytics and in‑app guidance for onboarding and feature adoption; overlaps on activation and insight use cases for product teams.