Radical Whale is a data workspace that lets you define datasets, attach AI-driven columns, and run enrichment or research workflows using your LLM keys, your APIs, and your tools.
You can:
• Create datasets with AI columns that compute values from instructions
• Build custom agents that call any API or MCP tool (Tavily, GitHub, Apollo, Diffbot, etc.)
• Mix text, datasets, and agent calls in TipTap-powered notebooks
• Run workloads in isolated per-workspace queues for predictable performance
I built this because tools like Attio, Notion, and Freckle hide the real agent calls behind credit systems and markup. By calling models and APIs directly, you can operate these workflows far more efficiently and transparently.
If you work with structured data, enrichment, or AI automation, I’d appreciate feedback on the approach.
Machinable is a passion project that I've just open sourced. Built as a custom BaaS to bootstrap apps, Machinable let's you create projects, then define API resources for that project using JSON schema. Automatically generate OpenAPI documentation, create web hooks to POST back to your own servers to handle business logic, and manage user sessions. I ultimately decided to open source it to see if others would find this useful as well.
You can:
• Create datasets with AI columns that compute values from instructions
• Build custom agents that call any API or MCP tool (Tavily, GitHub, Apollo, Diffbot, etc.)
• Mix text, datasets, and agent calls in TipTap-powered notebooks
• Run workloads in isolated per-workspace queues for predictable performance
I built this because tools like Attio, Notion, and Freckle hide the real agent calls behind credit systems and markup. By calling models and APIs directly, you can operate these workflows far more efficiently and transparently.
If you work with structured data, enrichment, or AI automation, I’d appreciate feedback on the approach.