Workshop: Building Production-Ready AI Pipelines: From Data to Deployment
Details
Why integrating AI into your stack shouldn't mean ripping it apart.
Most developers end up fighting their AI layer instead of building with it: context-switching between tools, rewiring workflows, stitching together integrations that were never meant to talk to each other.
RocketRide is an AI runtime built to fix that. Fast to integrate, easy to extend, scalable by design. Whether you work from an IDE or a terminal, RocketRide meets you there. You keep your workflow. We handle the AI layer.
In this workshop, we'll build a voice-driven coding agent that turns a voice memo into a working app. No forms, no prompts typed into a chat box. You describe what you need out loud, and the agent builds it.
What
RocketRide is an open-source AI runtime with 50+ nodes spanning LLM providers, agents, file system, shell execution, HTTP integrations, audio transcription, and more. You define what you want to build as a pipeline, and the runtime handles execution.
Today's project: a personal coding agent you talk to through a chat interface. You send it a voice memo describing the app you need, it transcribes your message, then autonomously scaffolds and runs the project. The whole thing is built with RocketRide nodes: an agent node with file system, shell, and audio transcription tools wired together as a pipeline.
The use case is the long tail of niche problems that are too specific for any product to ever solve. Instead of living with the workaround, you describe what you need and it gets built.
How
We open with a live build. A voice memo comes in through chat, audio transcription pulls the text, then file system and shell nodes scaffold and run a working app. In front of the room, start to finish.
Then it's your turn. Think of a tool you've always needed that's too niche to exist. Describe it out loud. Build it. At the end, you demo what you shipped.
Requirements
Bring a laptop with:
- Python and JavaScript installed
- An API key for your preferred LLM
- An IDE (VS Code, Cursor, or Windsurf recommended) with the RocketRide extension:
- VS Code Marketplace: https://marketplace.visualstudio.com/items?itemName=RocketRide.rocketride
- Open VSX: https://open-vsx.org/extension/RocketRide/rocketride
If you have a problem that's too specific for any product to ever solve, bring that too.
Lightning Talk:
- J Coleman