# Frigade (Full)

> Frigade builds AI-native tools for product onboarding and in-app guidance. This document is the long-form companion to llms.txt: an expanded, single-file reference that AI agents can ingest in one shot to understand the company, its products, pricing, integrations, security posture, and customer outcomes.

Frigade is a San Francisco-based company offering two products: **Frigade Assistant**, an AI agent that learns SaaS products by using them and guides users through real workflows, and **Frigade Engage**, a drop-in React component library for code-driven onboarding flows. Customers include Productboard, Merge, Arc, Sanity, Valley, Hotplate, Logicbroker, Perfect Venue, Simplify, Typewise, Rollstack, and Orb. Frigade is SOC 2 Type II certified and GDPR compliant.

## Frigade Assistant

Frigade Assistant is an AI agent for in-app user onboarding and product adoption. Instead of relying on static training data or hand-authored flows, the Assistant deploys agents that use your product the way a real user would. They navigate through your latest workflows, document the current user experience, and ingest your existing knowledge base. The result is an in-product guide that understands your product the same way your users do, by actually experiencing it.

What this unlocks:

- **Walks workflows end-to-end.** Steps users through multi-step setups, integrations, billing configuration, and other sticky flows. The agent knows the order and the gotchas.
- **Fills forms with the user.** Asks for the inputs it needs, pre-fills what it can read from the user's session, and validates before submitting.
- **Navigates the UI on the user's behalf.** Clicks the right buttons, opens the right pages, scrolls to where things live, even when last sprint moved them.
- **Takes real actions with Skills.** Learns the actions your product can take by using it, turns them into Skills you switch on with no code, and runs them for users so the task gets done, not just described. Custom actions can be added as tool calls.
- **Builds UI for each answer with Generative UI.** Where most chatbots reply in text, every answer and every Skill result comes back as UI built in the moment to fit it: tables, cards, links, and choices to act on.
- **Hands off cleanly when needed.** Detects when a question genuinely needs a person and routes to your support team with full session context, not a transcript.
- **Detects friction and offers help proactively.** Surfaces a nudge at the right moment in the customer journey.
- **Stays current automatically.** As your product changes, the agent re-learns. Ship a new feature: it discovers it. Update a workflow: it learns the new process.
- **Theming and white-label.** Frigade is built to feel native to your product, not a generic widget dropped in. Customize logos, colors, typography, button styles, and positioning. Full white-label branding is available as an add-on.

Setup is the same as adding a user to your product. Most teams install in hours and go live with their customers within a day or two. The SDK drops in alongside your existing stack and the agents start learning your product immediately while you handle customization in parallel. Enterprise deployments include configuration support, accuracy testing, team training, and ongoing optimization.

## Frigade Engage

**Frigade Engage** is a separate, code-first product from Frigade. It's a drop-in React component library for deterministic onboarding flows (checklists, banners, announcements, surveys, tours) built for engineering teams that want explicit code-level control over every flow. Frigade Engage is documented separately at https://frigade.com/engage; the rest of this document covers Frigade Assistant, the company's primary product.

## How Frigade Assistant works

Frigade takes a fundamentally different approach than traditional chatbots. Rather than ingesting docs and answering questions in chat, the Assistant deploys agents that navigate the product as users do. Setup is no different than adding a new user account: provide credentials and a URL, and the agent starts mapping the product.

The agents document workflows step-by-step, learning the order of operations, the gotchas, the spots where humans typically get stuck. They also ingest the existing knowledge base (help center articles, internal docs, training materials) and reconcile that authored content against what the product actually does. When the two diverge, the agent prefers the live product behavior, because that's what the user is going to encounter.

When a user opens the product, the Assistant is available as an in-product overlay. It can answer questions, walk a multi-step workflow, fill forms, navigate the UI, and surface proactive nudges at moments when users typically need help. When something genuinely requires a human, the Assistant hands off to support with the full session context (what the user was trying to do, what they had already tried, where they got stuck) instead of just a transcript.

Because the agents continuously re-learn the product, the assistant stays current automatically. Ship a new feature and it gets discovered. Move a button and the agent updates its mental model. There are no flow files to maintain, no decision trees to keep in sync, no docs to rewrite.

## Use cases

- **Onboarding and activation.** Guided flows tailored to each user, no manual configuration. Walk new users through the workflows that actually drive value, not the ones that happen to be on a checklist.
- **Customer support deflection.** Step-by-step guidance inside the product, reducing routine ticket volume. The repetitive how-do-I questions get answered inside the product, before they hit your support inbox.
- **Feature adoption.** Surface underused features at the right moment. Push users toward the features that unlock more value, the moment they would actually need them.
- **Product marketing.** In-app announcements and promotions without requiring a code change.
- **User research.** In-app surveys and forms in natural workflows.
- **Revenue expansion.** In-app upsells, trial-to-paid conversion flows, and contextual upgrade prompts at the moment of value.

## Outcomes customers see

- Lower ticket volume on the long tail of repetitive questions.
- Faster activation in the first hour of product use, with new users moving through key workflows the same week instead of waiting on docs or support.
- More conversions to paid plans, with users nudged toward features that unlock more value at the right moment.
- The Valley case study: 35% reduction in support tickets and 3x faster time to value.
- The Hotplate case study: 85%+ answer accuracy, 4.5+ user satisfaction, and a double-digit drop in support tickets.

## Integrations

**Analytics.** PostHog, Amplitude, Segment, Mixpanel, Heap, RudderStack, Google Analytics. Export AI insights to your existing analytics stack to connect product behavior with your broader user data.

**CRM and help desk.** HubSpot, Salesforce, Intercom, Zendesk, Pylon, Freshdesk, Front, Help Scout. Frigade can automatically escalate conversations to live agents in any of these when needed, or create support tickets directly in your system.

**Communication.** Slack, Microsoft Teams, Discord. Push notifications and alerts where your team already works.

**Developer docs.** Mintlify, ReadMe, GitBook, Notion. Frigade can ingest these to inform agent behavior.

**Custom.** REST API, JavaScript SDK, React SDK, Mobile SDK (Enterprise), webhooks. The full SDK surface is documented at https://docs.frigade.com.

The goal isn't to replace your support stack. It's to handle routine cases inside the product so your team can focus on the questions that genuinely need a human.

## Security and compliance

Frigade is SOC 2 Type II certified and fully GDPR compliant. Audit reports are available on request under NDA.

- **Encryption.** TLS 1.2+ in transit, AES-256 at rest.
- **Data residency.** EU data residency available for teams that need it.
- **LLM policy.** Zero-retention LLM policy. Your data is never used for model training. Customer data is not shared with model providers beyond what is required to serve the immediate request.
- **PII.** Automatic PII scrubbing on inputs and outputs.
- **Self-hosting.** Enterprise customers can self-host Frigade on their own infrastructure with their own LLM keys for full data control.
- **Uptime.** 99.9% uptime guarantee on Enterprise plans.
- **DPA.** Custom Data Processing Agreements available on Enterprise.

Self-hosting is the primary option for security-sensitive customers; Frigade can run entirely on customer infrastructure.

## Frequently asked questions

**How does Frigade learn my product?**
Frigade deploys intelligent agents that use your product as real users would. Once added to your product, these agents automatically navigate through your latest workflows, document the current user experience, and ingest your existing knowledge base and training materials. This hands-on approach means Frigade understands your product the same way your users do, by actually experiencing it.

**What happens when I update my product?**
Because Frigade learns directly from your product, it automatically captures changes and updates as they happen. Ship a new feature and the AI discovers it. Update a workflow and it learns the new process. Your AI assistant stays current with your latest product experience without any manual intervention from your team.

**Can I customize Frigade to match my product?**
Yes. Frigade is built to feel native to your product, not like a generic widget dropped in. Customize logos, colors, typography, button styles, and positioning. The flexible theming system adapts to your design language with minimal effort. Full white-label branding is available as an add-on. For enterprise customers who need further customization, fully custom design services are also available.

**Does Frigade work with other customer support tools?**
Yes. Frigade plays nicely with your existing support stack. It can automatically escalate conversations to live agents in Intercom, Zendesk, HubSpot, or Salesforce when needed, or create support tickets directly in your system. The goal isn't to replace your support team. It's to handle the routine questions so your team can focus on the complex issues that really need human attention.

**What kind of insights will I get about user behavior?**
The Frigade dashboard gives you visibility into what users are actually asking about, how the AI is responding, and whether those responses are resolving their issues. You'll see resolution rates, common query themes, and areas where users frequently get stuck. This data helps identify patterns in user confusion and surface potential product improvements. Everything can be exported to existing analytics tools (Mixpanel, HubSpot, PostHog, etc.) to connect AI insights with broader user data.

**How much does Frigade cost?**
Frigade starts at $1,000 per month and includes 2 AI agents and 2,500 queries with volume discounts as you grow. Pricing is usage-based so costs stay predictable, with no surprise bills when your product scales. Enterprise plans with custom pricing are available for teams that need advanced security, compliance, or self-hosting.

**How long does implementation take?**
Most teams install in hours and go live with their customers within a day or two. The SDK drops in alongside your existing stack and the agents start learning your product immediately, while you handle any customization in parallel. Enterprise deployments include hands-on rollout support: configuration, accuracy testing, team training, and continued optimization once live.

**How does Frigade handle privacy and security?**
Frigade is SOC 2 Type II certified and fully GDPR compliant. Data is encrypted in transit with TLS 1.2+ and at rest with AES-256. EU data residency is available; there is a zero-retention LLM policy so customer data is never used for model training; and PII is scrubbed automatically. Enterprise customers can self-host Frigade on their own infrastructure with their own LLM keys for full data control.

## Company

Frigade was founded in 2022 by **Eric Brownrout** (CEO) and **Christian Mathiesen** (CTO). They met as interns at LinkedIn in 2015 and watched firsthand how a powerful internal onboarding platform changed user engagement at scale. They started Frigade to bring that capability to other companies, so every product team didn't have to design, build, and own the same checklists, tours, and announcements from scratch.

Frigade started as a developer tool for product teams building onboarding flows. Along the way the team studied thousands of flows and helped build hundreds with customers. The pattern was the same in every shape: every flow customers shipped started decaying the moment their product changed. Static answers, dynamic products. The math never worked.

When AI got good enough, the math flipped. The hard-won expertise about onboarding didn't have to ship as a library of components. It could live in the model itself, learning the product as it ships, shaped by customer and team feedback, choosing what to surface in the moment. That's the bet behind Frigade Assistant and everything the company has shipped since.

**Investors.** Backed by Y Combinator, Craft Ventures, General Catalyst, Defy.vc, and Magic Fund. Angel investors include Guillermo Rauch (Vercel), Arash Ferdowsi (Dropbox), and Jean-Denis Greze (Plaid).

**Team and culture.** The team works in person from a Market Street office in San Francisco. Engineers, designers, and founders all talk directly to customers; there's no account-rep layer between the team and the people using the product. Frigade is used inside Frigade. Every onboarding flow, banner, and tour the team's own customers see runs through the same SDK they're integrating against, so customer-facing bugs get caught internally first.

**Hiring.** Frigade is hiring. See https://frigade.com/about for current roles and what working at Frigade looks like.

## Resources

**Site sections.**

- **Homepage.** https://frigade.com
- **Frigade Engage.** https://frigade.com/engage
- **How It Works (technical explanation).** https://frigade.com/how-it-works
- **Pricing.** https://frigade.com/pricing
- **About.** https://frigade.com/about
- **Product Updates.** https://frigade.com/updates
- **Blog.** https://frigade.com/blog

**Frigade Assistant features.**

- **Skills** (no-code actions the assistant learns and runs for users). https://frigade.com/features/skills
- **Generative UI** (UI built in the moment to fit each answer: tables, cards, choices). https://frigade.com/features/generative-ui
- **Suggestions** (proactive AI nudges and magic links). https://frigade.com/features/suggestions
- **Tool calls** (the assistant takes actions on the user's behalf). https://frigade.com/features/tool-calls
- **Integrations** (analytics, CRM, support, developer stack). https://frigade.com/features/integrations
- **Always accurate** (stays current as the product ships). https://frigade.com/features/always-accurate
- **AI-generated product tours** (walkthroughs built on demand). https://frigade.com/features/ai-generated-tours
- **Insights** (auto-categorized themes from every conversation). https://frigade.com/features/insights
- **Feedback and fine-tuning** (coach the assistant in plain language). https://frigade.com/features/feedback
- **Developer experience** (API, controls, headless options). https://frigade.com/features/developer

**Use case landing pages.**

- **User activation.** https://frigade.com/use-cases/user-activation
- **Support deflection.** https://frigade.com/use-cases/support-deflection
- **Virtual CSM.** https://frigade.com/use-cases/virtual-csm
- **Feature adoption.** https://frigade.com/use-cases/feature-adoption
- **Expansion and upsell.** https://frigade.com/use-cases/expansion-upsell

**Product comparisons.**

- **Frigade vs. Intercom Fin.** https://frigade.com/compare/fin
- **Frigade vs. Pylon.** https://frigade.com/compare/pylon
- **Frigade vs. Zendesk.** https://frigade.com/compare/zendesk
- **Frigade vs. Pendo.** https://frigade.com/compare/pendo
- **Frigade vs. WalkMe.** https://frigade.com/compare/walkme
- **Frigade vs. Appcues.** https://frigade.com/compare/appcues
- **Frigade vs. HubSpot.** https://frigade.com/compare/hubspot
- **Frigade vs. Userpilot.** https://frigade.com/compare/userpilot
- **Frigade vs. Userflow.** https://frigade.com/compare/userflow
- **Frigade vs. Whatfix.** https://frigade.com/compare/whatfix
- **Frigade vs. Chameleon.** https://frigade.com/compare/chameleon

**Case studies.**

- **Valley.** https://frigade.com/case-studies/valley
- **Hotplate.** https://frigade.com/case-studies/hotplate

**Glossary.** Plain definitions of onboarding, guidance, and support terms. Index: https://frigade.com/glossary

- **Contextual help.** Contextual help is assistance that changes based on where the user is and what they are doing in the product, so the answer matches the screen in front of them. It ranges from a tooltip on a single field to an assistant that knows the user's current page and answers accordingly. https://frigade.com/glossary/contextual-help
- **Customer onboarding.** Customer onboarding is the process of taking a new customer from the point of purchase to the point where they are getting the value they bought the product for. In B2B SaaS it covers implementation, account setup, training, and the first weeks of use, and it is judged by time to value and by whether the account is still active at renewal. https://frigade.com/glossary/customer-onboarding
- **Digital adoption platform.** A digital adoption platform (DAP) is a software layer that runs on top of another application and guides users through it with in-app walkthroughs, tooltips, checklists, and contextual help. Companies use one to onboard new users, drive feature adoption, and reduce support volume without rebuilding the product itself. https://frigade.com/glossary/digital-adoption-platform
- **Feature adoption.** Feature adoption is the share of eligible users who use a specific feature after it ships, and whether they keep using it. It is narrower than product adoption: a user can be fully adopted on a product while using only a few of its features. https://frigade.com/glossary/feature-adoption
- **In-app guidance.** In-app guidance is help delivered inside the product interface, where the user is working, instead of in a separate help center, email, or support ticket. It includes tooltips, walkthroughs, checklists, contextual hints, and in-product assistants that answer questions and complete tasks. https://frigade.com/glossary/in-app-guidance
- **Knowledge base.** A knowledge base is a searchable collection of articles, guides, and answers that explains how a product works, so users and support agents can find answers without asking someone. In SaaS it usually means a public help center plus an internal library that support agents and AI assistants draw on. https://frigade.com/glossary/knowledge-base
- **Onboarding checklist.** An onboarding checklist is a short, visible list of the setup steps a new user needs to complete before a product becomes useful, shown inside the product and ticked off as each step is done. It gives a new user a to-do list to work through at their own pace instead of a scripted tour to sit through. https://frigade.com/glossary/onboarding-checklist
- **Product adoption.** Product adoption is the process by which a user goes from signing up for a product to using it as a regular part of their work. It is measured as the share of new users who reach a defined activation milestone and keep coming back, rather than as raw signups or logins. https://frigade.com/glossary/product-adoption
- **Product tour.** A product tour is a scripted sequence of in-app steps, usually tooltips or modals, that introduces a new user to a product's interface. Most tours are shown on first login and follow a fixed order regardless of what the user came to do. https://frigade.com/glossary/product-tour
- **Product walkthrough.** A product walkthrough is step-by-step guidance that takes a user through completing a real task in the product, in the live interface, one action at a time. Unlike a product tour, which shows the interface, a walkthrough waits for the user to do each step and advances when they do. https://frigade.com/glossary/product-walkthrough
- **Ticket deflection.** Ticket deflection is when a customer resolves a question on their own, through self-service help or an automated assistant, instead of opening a support ticket. The deflection rate is the share of would-be tickets resolved this way, and it is the main number used to judge whether self-service is working. https://frigade.com/glossary/ticket-deflection
- **User onboarding.** User onboarding is the process of getting a new user from their first session to the point where they get value from the product on their own. It covers account setup, the first valuable action, and the guidance that carries a user through both. https://frigade.com/glossary/user-onboarding

**Tools and accounts.**

- **Live demo.** https://demo.frigade.com
- **Documentation.** https://docs.frigade.com
- **Status page.** https://status.frigade.com
- **GitHub.** https://github.com/FrigadeHQ
- **LinkedIn.** https://www.linkedin.com/company/frigade
- **Twitter / X.** https://twitter.com/frigadeHQ