Introducing Launchly: Live Ops Charts, Embedded Anywhere, in 60 Seconds

We built Launchly because we kept solving the same problem at every company we worked at: someone needed a live chart on a dashboard, a status page, or inside a customer portal — and suddenly it was a week-long project.

You know the drill. You pick a charting library. You write a backend endpoint to query your database. You figure out refresh logic. You deal with auth tokens. You fight with iframe CSS. And then six months later, someone asks for a second chart and you do the whole thing again.

What Launchly does

Launchly lets you go from "I need a chart" to a live, auto-refreshing web component in about 60 seconds:

1. Connect your data source — Postgres, CloudWatch, MySQL, ClickHouse, MongoDB, Snowflake, BigQuery, HTTP APIs, and more. Credentials are AES-256 encrypted at rest.

2. Describe what you want — Tell the AI "show me p95 API latency by endpoint for the last 6 hours" and it writes the Python script, picks the chart type, and runs a live preview.

3. Embed it — Copy two lines of HTML. The web component works everywhere: React, Vue, plain HTML, Notion, Retool, Confluence. Shadow DOM means zero CSS conflicts with your app.

That is it. No npm packages to install. No build step. No charting library to learn.

How it works under the hood

Every chart in Launchly is backed by a Python script with the signature def run(context) -> dict. The AI writes the first draft, and you can edit it in a Monaco editor if you want to tweak things.

Scripts run in an isolated AWS Lambda container with pandas, duckdb, and connector libraries pre-installed. Results are cached at the edge with a configurable TTL — your embeds read from cache on the hot path, not from Lambda on every page load.

The embed itself is a custom element with Shadow DOM:

html <script src="https://launchly.io/embed.js"></script> <launchly-chart token="eyJhbG..." />

Framework-agnostic. Style-isolated. Auto-refreshing. Signed JWT tokens scoped per chart with optional domain allowlists.

Who this is for

Developers and ops engineers at SaaS companies who need to embed a live metric somewhere and do not want to stand up Grafana or build a custom charting layer for it.

If you have ever:

• Wished you could just drop a CloudWatch graph into your internal dashboard without an AWS console login
• Needed to show a customer their API usage without building an analytics page from scratch
• Wanted a Postgres query result to auto-refresh on a status page

...then this is the tool we wish we had.

What is available today

• 14 data source connectors (CloudWatch, Postgres, MySQL, ClickHouse, MongoDB, Snowflake, BigQuery, GA4, Stripe, Datadog, PostHog, Linear, GitHub, HTTP/REST)
• AI-powered chart builder with iterative chat
• 10+ chart types (line, bar, area, pie, single-stat, table, gauge, heatmap, treemap, timeline)
• Interactive filters with cascading dependencies
• Dashboard composition with drag-and-drop layout
• Embeddable web components and hosted embed URLs
• CSV export and PNG screenshot

Get started

Launchly is free to start — no credit card required. Sign up at launchly.io/signup and create your first chart in under a minute.

We are a small team and we would love your feedback. Reach out at support@launchly.io or just reply to any email we send you.