AI AutomationInsights & Best Practices

Deep dives into workflow automation, AI agents, and building production-ready systems

Pearl Upgrade: Code-Based Sandbox Execution

December 15, 2025

PearlAI AgentsCode ExecutionSandboxContext GatheringBubbleFlow

Pearl Upgrade: Code-Based Sandbox Execution

We just gave our AI assistant the ability to write and execute code before building your workflows. Here's why that changes everything.

Selina Li

The Agentic Future is Code-First: Why We Built Bubble Lab on TypeScript, Not Hype

November 7, 2025

5 min read

TypeScriptAI AgentsMCPDeveloper ToolsCode Execution

The Agentic Future is Code-First: Why We Built Bubble Lab on TypeScript, Not Hype

Anthropic's recent article on code execution with MCP validates what we've known from day one: the most powerful way to build AI agents is with code. Here's why we built Bubble Lab on TypeScript, not protocols.

Selina Li

OpenAI Agent Builder vs. Bubble Lab: Why We're Excited, Not Worried

October 5, 2025

5 min read

OpenAIWorkflow AutomationAI AgentsOpen Source

OpenAI Agent Builder vs. Bubble Lab: Why We're Excited, Not Worried

OpenAI's Dev Day was a significant moment for workflow automation, and it validates what we've been building. Here's why we view Agent Builder as validation rather than competition, and why the future of workflow automation is open, transparent, and developer-owned.

Selina Li

The Industry Agrees: Code-First is the Future of AI Agents, and We've Been Building it Since Day One

October 1, 2025

5 min read

TypeScriptAI AgentsMCPDeveloper ToolsCode-First

The Industry Agrees: Code-First is the Future of AI Agents, and We've Been Building it Since Day One

Cloudflare's recent "Code Mode" blog post validates what we've known from the start: LLMs are far more powerful when they write code to interact with tools rather than calling them directly. Here's why this architectural shift matters—and why Bubble Lab has been code-native from day one.

Selina Li