Show HN: Oratio – Natural language programming that expands itself with AI (github.com)

Oratio is an open-source "natural language programming" language that maps plain-English (and Italian, with Spanish coming) instructions to executable operations. Demonstrations show simple NL scripts like "Load sales.csv. Filter rows where amount > 1000. Calculate the average. Create a bar chart." being parsed into concrete operations (io.read_csv, math.mean) and runnable Python/pandas code, with a runtime/CLI that executes the generated workflow. The project ships core runtime, basic operations, an IDE/CLI, cloud integration, a plugin system, and a marketplace roadmap; it's MIT-licensed and invites contributions for new operations, connectors, translations and ML/robotics integrations. For the AI/ML community, Oratio is significant because it lowers the barrier to programmatic data analysis and workflow automation—enabling non-programmers to express tasks in natural language and enabling rapid prototyping of data/ML pipelines. Technically, it centers on semantic parsing that maps intent to a library of typed operations and an execution engine, which makes extensibility via connectors and plugins straightforward. Important implications include opportunities to standardize NL→code semantics, build reusable ML operations and MLOps connectors, and research correctness, reproducibility, and safety (sandboxing, data handling, and ambiguous intent). Oratio’s open-source model lets researchers and engineers contribute operations, benchmarks, and safety audits as the platform expands into databases, APIs, ML, and IoT control.