# VerLake — Full Product Description

> Reliability monitoring platform that runs Playwright-based user journey tests on a schedule.

## Overview

VerLake is a SaaS platform for synthetic monitoring of real business workflows. Unlike traditional uptime monitoring that checks whether a URL returns HTTP 200, VerLake executes actual multi-step browser sessions using Playwright to simulate what real users do — logging in, searching, checking out, navigating dashboards, completing forms.

When a monitored user journey fails, VerLake captures screenshots, video recordings, browser console output, and network request logs. An AI-powered analysis engine generates a root cause summary, making it fast to diagnose and fix production issues before users report them.

## Core Concepts

### User Journeys
A user journey is a Playwright test script that simulates a real user workflow. Scripts are written in JavaScript/TypeScript using the @playwright/test framework. VerLake provides a built-in code editor with an AI assistant (powered by Claude) that can generate scripts from plain-English descriptions, modify existing scripts, and debug failures.

Example user journey: "Agent logs in → dashboard loads → toggles availability → places outbound call → verifies call connected."

### Monitors
A monitor is a scheduled execution of a user journey. You configure:
- Which user journey to run
- How often (every 1 minute, 5 minutes, 15 minutes, 30 minutes, or 1 hour)
- Which environment to run against (production, staging, etc.)
- Failure threshold (consecutive failures before marking the monitor as failing)
- Timeout (maximum execution time per run)

Monitors run automatically on schedule and can also be triggered manually.

### Runs
A run is a single execution of a monitor. Each run records:
- Status: PASSED, FAILED, or TIMEOUT
- Duration in milliseconds
- Step-by-step breakdown with timing
- Error messages (if failed)
- Screenshots captured during execution
- Video recording of the browser session
- Browser console logs
- Network request logs
- AI-generated failure summary (for failed runs)

### Environments
Environments allow the same user journey script to run against different targets without modification. Each environment defines key-value variables (e.g., APP_URL, USERNAME, PASSWORD) that are injected into the Playwright script at runtime via process.env.

### Projects
Projects organize user journeys and monitors into logical groups — typically one per product or service being monitored.

### Status Pages
VerLake generates public status pages that display real-time uptime data from monitor results. Teams can share these with customers to communicate service health transparently.

## How It Works

1. **Author**: Write a Playwright script in the built-in editor. Use the AI assistant to generate scripts from natural language, or write them manually. Test scripts immediately with one-click test runs that return screenshots and video.

2. **Monitor**: Create a monitor that runs the script on a schedule. Pick the environment, interval, and failure threshold. VerLake handles scheduling, execution, and artifact storage.

3. **Debug**: When a run fails, inspect the full debug output — screenshots at each step, video replay, console errors, network requests, and an AI-generated root cause summary. Fix the issue before your users notice.

4. **Report**: Track reliability over time with dashboards showing uptime percentages, failure trends, and execution durations. Share public status pages with customers.

## Technical Architecture

- **Backend**: Java 17 / Spring Boot running on AWS (Elastic Beanstalk)
- **Browser automation**: Playwright with pooled Chromium browser instances
- **Database**: PostgreSQL (AWS RDS)
- **Frontend**: React + Vite single-page application
- **AI**: Claude (Anthropic) for script assistance and failure analysis
- **Multi-tenancy**: Organization-based isolation with JWT authentication and Google OAuth

## Pricing Model

VerLake uses usage-based pricing measured in **seconds of user journey execution time**. You pay only for actual browser automation time consumed by your monitors.

### Free Tier
- 1,000 execution seconds per month
- Up to 5 monitors
- 1 project
- Screenshots on failure
- AI script assistant

### Growth Plan — $0.005 per second
- Pay-as-you-go execution
- Unlimited monitors and projects
- Screenshots and video on failure
- AI script assistant and failure summaries
- Multi-environment support
- Team collaboration
- Public status pages

### Enterprise — Custom Pricing
- Volume discounts
- Dedicated infrastructure
- SSO/SAML
- Custom data retention
- SLA guarantees

### Pricing Example
A monitor that runs a 10-second checkout flow every 5 minutes:
- 10 seconds × 12 runs/hour × 24 hours × 30 days = 86,400 seconds/month
- Cost: 86,400 × $0.005 = $432/month

A simpler 3-second login check every 15 minutes:
- 3 seconds × 4 runs/hour × 24 hours × 30 days = 8,640 seconds/month
- Cost: 8,640 × $0.005 = $43.20/month

## Comparison to Alternatives

| Feature | VerLake | Checkly | Pingdom | UptimeRobot |
|---------|---------|---------|---------|-------------|
| Real browser tests | Yes (Playwright) | Yes (Playwright) | Limited | No |
| AI script assistant | Yes (Claude) | No | No | No |
| AI failure analysis | Yes | No | No | No |
| Multi-step workflows | Yes | Yes | No | No |
| Video on failure | Yes | No | No | No |
| Usage-based pricing | Per second | Per check run | Per check | Per monitor |
| Status pages | Included | Separate product | Add-on | Basic |

## Use Cases

- **E-commerce**: Monitor checkout, search, and account flows to catch breaks before customers abandon carts.
- **SaaS products**: Verify login, onboarding, core features, and integrations work across every deploy.
- **Contact centers**: Test agent login, call placement, and dashboard loading to maintain service levels.
- **Financial services**: Validate transaction flows, account access, and compliance-critical workflows.
- **Healthcare**: Monitor patient portal access, appointment booking, and prescription refill flows.

## Links

- Website: https://verlake.com
- Application: https://app.verlake.com
- Sign up: https://app.verlake.com/signup
- Contact sales: sales@verlake.com
- General inquiries: hello@verlake.com
- Support: support@verlake.com