Show HN: Telize – YAML workflows for shell, Python, and LLM steps (github.com)

Telize has launched a low-code framework that enables developers to create reproducible, structured AI workflows using YAML. This tool allows users to seamlessly integrate various components—such as LLM calls, shell commands, and Python functions—within a single workflow file. With features like upfront configuration validation, live progress tracking, and the ability to compose steps using Jinja templating, Telize simplifies the complex process of chaining together multiple APIs and scripts, which has traditionally resulted in unwieldy codebases. The significance of Telize lies in its potential to address a prevalent challenge in AI engineering: the unpredictability and complexity associated with existing frameworks. By providing a deterministic structure (using YAML) while allowing AI to handle creative tasks, it offers a more manageable solution for developers. This not only facilitates rapid prototyping—allowing users to build agentic workflows in minutes—but also mitigates the costs associated with running erroneous configurations, making it particularly appealing for enterprise and production environments where reliability is paramount. Additionally, the framework's local-first approach and adherence to DevOps principles will likely resonate with those already familiar with tools like Ansible and Docker Compose.