Rectify
An all-in-one operations platform for SaaS founders, bundling session replay, uptime monitoring, an AI support inbox, code scanning, roadmap and changelogs into one workspace driven by conversation, with bring-your-own-key AI models charged at cost.
What is Rectify?
Rectify is an all-in-one operations platform aimed at SaaS founders, including those without a technical background. Rather than stitching together a session replay tool, a feedback widget, an uptime monitor, a code scanner, a roadmap board and a changelog generator, it gathers those functions into one workspace and lets you drive them by conversation.
That conversational layer is an in-house AI agent named Quanta, and the publisher is blunt about its role: it executes rather than advises. You can ask it to find sessions where users hit errors, rage-clicked or dropped off, with no filters to configure; to draft support replies carrying the full conversation context and escalate to a human when that is the right call; or to pull a code scan report, check uptime, generate a changelog and move a roadmap card. Because it reads across the whole workspace, it can tie a support ticket to a session replay, or trace a bug from a code scan through to a roadmap card.
The feature surface is wide. Uptime monitoring covers HTTP/HTTPS, TCP, DNS and ICMP, with alerts to Slack, Discord, email or webhooks, branded status pages, incident tracking with automatic MTTR calculation, and SSL certificate expiry checks. Code Scan runs static analysis on GitHub repositories and returns bugs, security vulnerabilities and code smells with A-E ratings for maintainability, reliability and security, pinpointed to file, line and character offset with fix recommendations. Bug reporting captures annotated screenshots, Loom recordings, console logs, network requests and browser details in a single click. A Chrome extension adds annotated session recording, a PRD generator that outputs structured JSON, and reusable tagged code snippets. AgentPulse goes further still, letting an agent create and orchestrate sub-agents, install skills, set cron schedules and report usage costs.
An MCP server connects Claude, Cursor or any compatible client, either as a remote HTTP server over OAuth 2.0 with PKCE or locally through npx, exposing read and write tools across every module. Pricing rests on a bring your own key model: you supply your own OpenAI, Anthropic or Google credentials and pay the provider directly, with no markup.
What it does
- Replay user sessions and automatically surface rage clicks, errors and drop-off points
- Monitor infrastructure over HTTP/HTTPS, TCP, DNS and ICMP, alerting to Slack, Discord, email or webhooks
- Scan a GitHub repository for bugs, security vulnerabilities and code smells, rated A to E
- Collect bug reports carrying annotated screenshots, Loom recordings, console logs and network requests
- Handle customer conversations from a single AI-assisted support inbox with three escalation modes
- Generate release notes from commits and run a public roadmap with user voting
- Drive the entire platform in plain language through the Quanta agent or any MCP client
When to use Rectify / When not to
A quick filter to help you decide if Rectify is the right fit.
When to use Rectify
- Solo founders and small SaaS teams of up to 25 people, which is where the plan seat limits stop
- Teams shipping with AI coding assistants such as Cursor, Claude, Lovable, Bolt, v0 or Replit, whose debugging still happens the old way
- Product teams currently paying for seven to ten separate tools and looking to consolidate the spend into one subscription
- Anyone who needs to link a support ticket, a session replay, a code scan and a roadmap card without switching tabs
- Users who already hold their own OpenAI, Anthropic or Google API keys and want to avoid a reseller markup on AI usage
When not to use Rectify
- Organisations whose procurement requires a signed Data Processing Agreement, since none is published or offered on the site
- Buyers who need recognised security certifications: no SOC 2, ISO 27001 or HIPAA attestation appears anywhere on the site
- Teams that must contractually exclude their data from AI model training, as no opt-out mechanism is documented
- Anyone needing a native mobile app, the product being a web platform plus a Chrome extension only
- Non-English-speaking teams, as the interface and all published material are in English with no language selector
How to use Rectify
A typical end-to-end flow, from setup to results.
- Create an account on app.rectify.so, either on the permanent free tier or through the 14-day trial, which asks for no credit card
- Install the Rectify widget in your own application so that users can send feedback and bug reports with technical context attached
- Add your own OpenAI, Anthropic or Google API keys in settings to activate the bring-your-own-key AI models
- Connect a GitHub repository to trigger on-demand static analysis scans
- Create your first monitor, choosing the protocol and the alert channels you want to be notified on
- Set up a branded status page and let incident tracking record MTTR automatically
- Install the Chrome extension from the Chrome Web Store for annotated session recording, PRD generation and code snippets
- Connect an MCP client such as Claude or Cursor: either point it at the remote MCP URL shown in settings and complete the OAuth flow, or paste the npx configuration block with a token generated in settings
- Open the roadmap and changelog modules to collect feature votes and publish release notes
- Build workflows that trigger on feedback, roadmap, code scan, monitoring, inbox or changelog events, and review the run history
Pros & Cons
Pros
- Consolidates a stack the publisher costs at 400 to 1,500 USD a month into a single subscription of 29 to 99 USD
- Bring-your-own-key AI with no markup: usage is billed by the provider directly, and a cost cap feature is advertised
- A permanent free tier alongside a 14-day trial on every paid plan, neither of which requires a credit card
- The MCP server exposes write operations, not just reads, so a connected agent can act rather than merely report
- Cross-module correlation means a ticket, a session, a scan and a roadmap card live in the same workspace
- Pricing is published in full, with a per-plan quota comparison table rather than a request-a-quote wall
- The privacy policy names every processor individually along with its place of processing
Cons
- No Data Processing Agreement is published or offered, which will block many regulated or enterprise buyers
- No security certification is displayed anywhere: no SOC 2, ISO 27001 or HIPAA attestation
- The publisher takes no published position on whether customer data trains AI models, and documents no opt-out
- No native mobile application, and the interface is English-only with no language selector
- A three-person team operating from a shared Covent Garden registered address raises a service-continuity question
- MCP documentation is published as a blog article rather than in a dedicated documentation space
- Traction figures are inconsistent between pages, the homepage claiming 1K+ products and the About page 800+ businesses
Pricing & Plans
A permanent free plan is available under the name Idea, limited to one user, one project, ten session replays and a single monitor. The cheapest paid entry point is the Build plan at 29.00 USD per month, followed by Launch at 59.00 USD and Accelerate at 99.00 USD. An annual commitment is advertised at a 20% saving, and a one-off Believer Plan grants lifetime access to the Accelerate tier for 699.00 USD. Enterprise pricing is quoted on request. Every paid plan opens with a 14-day free trial requiring no credit card. AI usage beyond the one-off credits is billed by the model provider directly under the bring-your-own-key model, not by Rectify.
- Free - 1 user
- 1 project
- 10 session replays
- 1 monitor
- 10 support responses
- 1 PRD
- no AI credits
- Deepseek and Gemini models only
- 29 USD/month - 3 users
- 3 projects
- 100 session replays
- 3 monitors
- 10 code scans
- 10 PRDs
- 5 USD of one-time AI credits
- 59 USD/month - 10 users
- 10 projects
- 1
- 000 session replays
- 10 monitors
- 25 code scans
- 25 PRDs
- 10 USD of one-time AI credits
- 99 USD/month - 25 users
- 25 projects
- 5
- 000 session replays
- 25 monitors
- 100 code scans
- 100 PRDs
- 15 USD of one-time AI credits
- 699.00 USD once - lifetime access to the Accelerate plan
- custom pricing - tailored solutions for teams at scale
- arranged through the contact page
Data, GDPR & hosting
A consolidated view of how Rectify handles your data.
GDPR overview
The privacy policy carries a dedicated GDPR section listing the full set of rights: withdrawing consent, objecting to processing, access, verification and rectification, restriction, erasure, portability and lodging a complaint with a supervisory authority. Requests are free and answered within one month, and objection to direct marketing requires no justification. Legal bases are enumerated, and users may ask about the basis for transfers outside the European Union. A minimum age of 16 is stated, with deletion available on request to privacy@rectify.so. Every processor is named alongside its place of processing. Two gaps are worth noting: no Article 27 EU representative is designated, the publisher being British, and no Data Processing Agreement is published or offered anywhere on the site.
Who owns the data?
The data controller is Replicate Technologies LTD, a company based at 71-75 Shelton Street, Covent Garden, London WC2H 9JQ, reachable at support@rectify.so and, for privacy matters, at privacy@rectify.so. Its privacy policy, generated through iubenda and last updated on 23 April 2026, names each processor it relies on and the country in which that processor operates. Site content and intellectual property remain the property of the publisher or its licensors under the terms of service. Where an AI assistant is connected over MCP, Rectify states that it neither stores nor exports the resulting chat history, and that API keys and credentials are never exposed through that integration.
Reuse rights
Users cannot freely reuse material published on the site. The terms of service reserve all rights: without prior written permission, no one may copy, download, share, modify, translate, transform, publish, transmit, sell, sublicense, edit or transfer the content, nor create derivative works from it, and the same restriction extends to anyone acting through a user's device. Downloading or sharing is permitted only where the site states so explicitly, for personal and non-commercial use, and provided copyright attributions are kept intact. Statutory exceptions to copyright are unaffected. Separately, users may reach their own data through the API, but the publisher disclaims any liability for damage arising from that use or from third-party products accessing data through it.
Data retention & training
Hosting summary
The publisher is established in the United Kingdom and the privacy policy names each processor with its declared place of processing. Hosting and backend infrastructure run on Amazon Web Services, with the United Kingdom given as the place of processing. Payments go through Stripe Payments Ltd, also in the United Kingdom. Analytics and authentication rely on Google Analytics 4 and Google OAuth, both operated by Google Ireland Limited and processed in Ireland. Traffic optimisation and distribution use Cloudflare, processed in the United States, as does the Encharge email service. Where an AI assistant is connected over MCP, Anthropic acts as a processor in the United States on the basis of user consent, with traffic encrypted using TLS 1.2 or higher. The policy notes that depending on a user's location, transfers may move data to another country, and points to the per-service sections for detail. No hosting region is claimed beyond these named countries.
Things to keep in mind
Risks and trade-offs to weigh before adopting Rectify.
- No Data Processing Agreement is published, so any professional use involving personal data rests on an unresolved contractual footing
- The widget records your own end users' console logs, network requests and browser details: you must inform them and update your own privacy notice accordingly
- Under the bring-your-own-key model your data travels to the AI provider you choose and falls under that provider's policy, not Rectify's
- No published position on training AI models with customer data, and no documented opt-out, leaves a real governance blind spot
- MCP access tokens stay valid for a year unless revoked, so a forgotten connection remains an open door onto your workspace
- Delegating operations to an agent that acts rather than advises can erode the habit of reading incidents first-hand and dull the team's own diagnostic reflexes
- Traction figures are unaudited and contradict each other between pages, so they should carry no weight in a purchasing decision
Setup & Integrations
Technical difficulty
Low to moderate, and deliberately so: the publisher targets non-technical founders. Two customers describe the widget as quick and straightforward to install. The Chrome extension comes from the Chrome Web Store, and a remote MCP connection is an OAuth flow with no local install. Three steps ask a little more: connecting a GitHub repository for code scanning, creating monitors by protocol and alert channel, and adding your own AI provider keys. The local MCP option requires pasting a short npx configuration block with a token, which the vendor sums up as one line of config.
Deployment
Integrations
Behind Rectify
Social
Resources
All the official URLs gathered for verification and reference.
Alternatives
Tools that compete with or complement Rectify.
Frequently asked questions
What is Rectify?
What is Quanta?
Is there a free plan, and do I need a credit card?
How does the bring-your-own-key model work?
Can I connect Rectify to Claude or Cursor?
What technical data does the bug reporting widget capture?
Is there a minimum age to use Rectify?
Is a Data Processing Agreement available?
Is there a mobile app?
Who publishes Rectify?
Should you pick Rectify?
Rectify makes a coherent bet: founders shipping fast with AI coding assistants end up with an operations stack of seven to ten disconnected tools, and the switching cost between them is where their time goes. Folding session replay, monitoring, support, code scanning, roadmap and changelog into one workspace, then putting a conversational agent on top of all of it, is a genuine answer to that problem rather than a repackaging of one. The bring-your-own-key pricing is the most distinctive commercial choice: at 29 to 99 USD a month against the 400 to 1,500 USD the publisher attributes to an equivalent stack, and with AI usage billed at cost by the provider, the value argument is easy to follow. The MCP server deserves particular attention, since it exposes write operations and not merely reads, which makes a connected agent an operator rather than a commentator.
The reservations are just as clear, and they are mostly about assurance rather than capability. No Data Processing Agreement is published, no security certification is displayed, and nothing is said about whether customer data trains AI models or how to exclude it. For a product that captures console logs, network requests and browser details from end users, those silences matter, and any regulated buyer will need to resolve them before signing. The publisher is a three-person team operating from a shared registered address, which is worth weighing for a tool that would sit at the centre of a product's operations.
The practical conclusion is favourable but conditional. A permanent free tier and a 14-day trial without a credit card make the platform genuinely cheap to evaluate, so the sensible route is to test it against a real workload and treat the contractual and data-governance questions as a separate conversation with the publisher.
- Choosing a selection results in a full page refresh.
- Opens in a new window.