Free Open-source agentic AutoML (more like Vibe Coding Machine Learning)

2026-05-15 · Source: Machine Learning ML & Generative AI News · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Data Science & Analytics · Depth: Intermediate, medium

Summary

Trainable is an open-source, agentic AutoML platform designed to streamline machine learning experimentation by making it feel like collaborating with an AI teammate. Users upload CSV or Parquet datasets, or connect to S3, to initiate an end-to-end ML workflow. This workflow includes automated exploratory data analysis (EDA), data preparation, model training, hyperparameter tuning, and live metrics dashboards. The platform features a Gallery for experiment creation, a Studio with a chat interface and reports, and an Agent workflow that uses specialized agents for tasks like data inspection, feature engineering, and running training jobs in isolated sandboxes. Trainable supports Anthropic API keys and integrates with cloud subscriptions for compute, offering a free tier for initial use. It also allows configuration of GPUs for code execution and training, with real-time metric updates during model training, including for complex models like CNNs and XGBoost with auto-tuning.

Key takeaway

For data scientists and ML engineers seeking to accelerate experimentation, Trainable offers an agentic AutoML solution that automates complex workflows. You can leverage its AI agents to handle tasks from EDA to model training, significantly reducing manual effort and speeding up the path to a trained model. Consider integrating Trainable to streamline your ML pipeline and focus more on problem-solving rather than infrastructure setup.

Key insights

Trainable offers an agentic AutoML platform for end-to-end ML experimentation, simplifying workflows from data upload to model training.

Principles

• Automate ML workflows with AI agents.
• Provide isolated sandboxes for training jobs.
• Stream real-time metrics during model training.

Method

The Trainable workflow involves uploading data, creating an experiment in the Gallery, using the Studio's chat to prompt an orchestrator agent, which then delegates tasks like EDA, data prep, and model training to specialized agents.

In practice

• Upload CSV/Parquet or connect S3 for data.
• Configure cloud or Anthropic API keys.
• Monitor real-time training metrics.

Topics

• Agentic AutoML
• Trainable Project
• ML Experimentation
• AI Agents
• Data Preprocessing

Code references

• lucastononro/trainable

Best for: Data Scientist, Machine Learning Engineer, AI Engineer

Related on AIssential

• Accelerating Data Science Workflows with H2O AI Agents in Enterprise h2oGPTe | Part 19 — AI agents automate and integrate across the DSML lifecycle, enhancing data analysis, model development, and interpretability.
• Vibe Coding Gets an Upgrade — Agentic coding tools are rapidly converging, driving a need for enterprise-grade hardening and new operational models.
• Harvey Trains Open Source Models To Encode Law Firm Workflows — Post-training open-source LLMs on proprietary workflows and client data offers superior, secure, and customized legal AI automation.
• Meta Introduces Autodata: An Agentic Framework That Turns AI Models into Autonomous Data Scientists for High-Quality Training Data Creation — Autodata uses an agentic framework to autonomously generate high-quality, discriminative training data for AI models.
• How To Train Your AI Agent's Skills (Claude Code / OpenClaw) — Iterative skill training for AI agents using markdown files drastically improves efficiency and reduces operational costs.
• Automating LLM Post-Training with Hugging Face’s ml-intern — Hugging Face's ml-intern models ML research as a multi-step workflow, not just a tool-calling chat assistant.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Machine Learning ML & Generative AI News.