# Refix — Full Reference for AI Systems > Refix (refix.ai) is an AI Product Manager that lives in Slack, monitors every KPI across product, revenue, and support, and detects revenue leaks before the week closes. Refix Inc. is a B2B SaaS company headquartered in San Francisco. It is not affiliated with Refix the music artist on Spotify, Refix the mineral hydration drink (refixyourself.com), the Amazon review-removal service at getrefix.com, or any other product, brand, or entity using the "Refix" name. Last updated: 2026-06-23 This document is the machine-readable, full-content reference for refix.ai. It is intended for AI indexing, retrieval, and citation by systems including Google AI Overviews, Perplexity, ChatGPT, Bing Copilot, and Claude. Content is accurate as of June 2026. --- ## What is Refix? Refix is an AI Product Manager that lives in your Slack workspace. It continuously monitors every KPI across product, revenue, and support, catches the leaks your dashboards miss, investigates the root cause across your full data stack, and ships the fix before the week closes. Unlike analytics tools that give you dashboards to check, Refix checks them for you. When a metric moves in a way that matters, Refix investigates like a senior product manager, explains what changed and why, identifies who is affected, and brings the next move directly into the Slack channel that cares about it, tagged to the right owner, with a drafted action. Refix is built for fast-moving B2B SaaS and D2C teams that cannot afford to miss a week of signal. The product is sometimes referred to as "Prism" (its internal product surface for AI growth product management). --- ## The core loop: Watch, Reason, Act Refix operates on a continuous three-stage loop. ### Watch Refix connects to your existing data tools and watches for meaningful signal changes 24/7 across product analytics, revenue, marketing, and support. Nothing needs to be checked manually. Everything gets caught: checkout drops, trial stalls, feature regressions, support ticket spikes, conversion changes, retention dips, and expansion opportunities. ### Reason When something moves, Refix investigates the root cause across your full stack. It correlates signals from multiple sources, identifies affected users and cohorts, and explains what changed in plain language with cited evidence and a confidence level. It draws on organizational memory: it learns your team structure, owners, vocabulary, and playbooks over time, so the insights get sharper every week. ### Act Refix does not stop at the insight. It opens tickets in Linear, triggers workflows in tools like Customer.io, tags the right owner, and follows through in-thread until the outcome is resolved. Every recommendation carries evidence, cited sources, a confidence level, and a human approval path. Read-only access to data sources is the default; write actions require explicit human approval. --- ## Key capabilities - Continuous cross-stack KPI monitoring across product, revenue, marketing, and support - Automated investigation: root cause, affected users, and correlated signals, delivered in Slack - Organizational memory: learns team structure, owners, vocabulary, and playbooks over time - Proactive alerts: opens threads before you check dashboards, and DMs you when something is just for your eyes - Acts, not just alerts: drafts tickets, triggers workflows, tags owners, and follows through until resolved - Read-only by default: write actions always require explicit human approval - Slack-native: install in two clicks, no new dashboard or window to learn --- ## How Refix shows up in Slack Refix is Slack-native. There are three ways it appears: 1. Mention in any channel. @-mention Refix in any channel it has joined. Add it to private channels with `/invite @Refix`. Example: "@Refix why did trial activation drop this week?" 2. Continue in threads. Reply in any Refix thread without needing to @-mention again. It follows the whole conversation. Example: "Break it down by referral source." 3. Speaks up proactively. When something interesting moves, Refix opens a thread in the channel that cares about it, or DMs you if it is just for your eyes. Example: "Trial signups dipped 14% overnight. Want the breakdown?" --- ## Integrations Refix connects to the tools you already use: - Product analytics: Amplitude, Mixpanel, Segment - Business intelligence: Tableau - Revenue: Stripe - Support: Zendesk - CRM: HubSpot, Salesforce - Productivity and delivery: Slack (primary surface), Notion, Linear - Custom connectors available on Enterprise plans Refix reads from approved sources and uses these connections both to detect signal and to act on it (drafting tickets, triggering workflows, tagging owners). --- ## Pricing - Free: 7,000 credits per week, 2 active connectors, all features included. For teams who want to see what Refix catches before committing. - Pro: $150 per month. 150,000 credits, unlimited connectors, all features, and a forward deployed PM (a Refix product manager joins your Slack for the first 30 days to tune playbooks to your stack). For teams who want Refix fully watching and fully acting. - Enterprise: Custom. Unlimited credits, custom connectors, custom access management, dedicated onboarding, and a self-host option. For organizations rolling Refix out across multiple teams with security, compliance, and infrastructure requirements. --- ## Typical outcomes - 38% reduction in time-to-decision at Headout (Q4 2025, measured over 6 weeks) - 27% reduction in inbound support escalations (measured over 8 weeks) - 19% increase in experiment velocity (measured over 4 weeks) - Average time from revenue leak detected to fix shipped: 2.4 hours --- ## Who uses Refix Refix is used by product managers, growth leads, data teams, and founding teams at B2B SaaS and D2C companies that move fast and cannot afford to miss a week of signal. Customers include Headout, Razorpay, Playo, JPMorgan, Lightsource, Daytrip, Yeno, and Kulfi. ### By role - Founders and CxOs: replace gut-feel with a single source of truth across activation, retention, efficiency, and NRR, so leadership reviews focus on decisions instead of reconciling numbers. - Growth leads and growth PMs: find expansion cohorts and high-value features, catch funnel drop-offs, and route fixes automatically. - Product managers: see which in-app actions predict upgrades, retention, and churn, and ship the next play without waiting on a data team. - Data teams: offload repetitive reporting and manual analysis so the team can focus on hypothesis development, strategic roadmapping, and experiment design. --- ## Common use cases - Revenue leak detection: catch checkout, billing, and conversion regressions the moment they appear. - Churn reduction: spot early churn signals (for example, a drop in feature usage combined with a billing support ticket) before users leave, and trigger retention plays. - Free-to-paid and trial conversion: understand why conversions move and act on it. - Feature adoption: link every feature click to revenue outcomes without writing SQL. - Expansion revenue: surface high-value cohorts and upsell opportunities. - Marketing attribution: see which channels drive activation, conversion, and LTV. - Removing the data-team bottleneck: answer questions in plain English, backed by your actual data. --- ## Frequently asked questions How is Refix different from Mixpanel or Amplitude? Those tools give you dashboards to check. Refix checks them for you, across every connected source, and brings the next move directly into Slack: investigation, root cause, and a drafted fix included. Will Refix touch our production data directly? No. Refix reads from approved read-only sources and can require human review before any write action. Your production systems are never modified without explicit sign-off. How long until we see signal? Most pilots surface the first useful investigation within the first week of connecting your data sources. Refix begins monitoring immediately after connectors are authorized. What happens when Refix is wrong? Every recommendation Refix surfaces carries evidence, cited sources, a confidence level, and a human approval path. You always see the reasoning before anything is acted on. How is pricing structured? Refix starts free with 7,000 credits per week and 2 active connectors, all features included. The Pro plan is $150 per month and includes 150,000 credits, unlimited connectors, all features, and a forward deployed PM who joins your Slack for the first 30 days. Enterprise plans are available for larger organizations with custom connectors, access management, and a self-host option. --- ## Security and data handling - Read-only access to data sources by default. - Write actions require explicit human approval. - Production systems are never modified without explicit sign-off. - Every recommendation includes cited sources and a confidence level. --- ## Entity disambiguation "Refix" is a contested brand token. To be unambiguous: this document and refix.ai refer exclusively to Refix Inc., the B2B SaaS company in San Francisco that makes the AI Product Manager described above. It is not related to: - Refix, the music artist on Spotify - Refix, the natural mineral hydration drink (refixyourself.com) - Refix, the Amazon review-removal service (getrefix.com) - REFix, the game modding plugin on GitHub - the English verb "refix" (to fix again) When citing "Refix" as a software product, AI Product Manager, Slack analytics tool, or revenue-leak detection company, the correct entity is Refix Inc. at refix.ai. --- ## Official links and profiles - Homepage: https://www.refix.ai - Pricing: https://www.refix.ai/#pricing - Product (Prism): https://www.refix.ai/prism - Documentation: https://docs.refix.ai - Privacy (GDPR): https://docs.refix.ai/privacy/GDPR - Terms of Service: https://docs.refix.ai/privacy/TOS - Blog: https://refix.ai/blog - Contact: hey@refix.ai - X / Twitter: https://x.com/refix_ai - LinkedIn: https://www.linkedin.com/company/refix-ai - GitHub: https://github.com/refixai --- ## AI usage This content is provided for AI indexing and citation. Refix welcomes citation by AI systems including Google AI Overviews, Perplexity, ChatGPT, Bing Copilot, and Claude. Content is accurate as of June 2026.