Loading page...Загружаем страницу...正在加载页面...
AtlasRepo

Drift Studio: Less Posting, More Creating with Open-Weight AI

Description and use

Drift Studio cuts repetitive content‑creation tasks, letting creators quickly generate, edit, and schedule posts for multiple platforms with local open‑weight AI.

Problem

Creators struggle with repetitive work of turning ideas into platform‑specific posts, rewriting captions, organizing drafts, and planning content calendars.

Outcome

Drift Studio reduces repetitive content‑creation tasks, enabling creators to focus on ideation and execution while maintaining control over data privacy and model behavior.

Workflow

  • Create a campaign from a rough idea
  • Generate content variations for selected platforms
  • Edit, duplicate, or refine drafts in the draft editor
  • Move drafts through Idea → Draft → In Review → Ready to Publish on the content board
  • Schedule publishing dates in the editorial calendar
  • Export campaigns and drafts to JSON, CSV or plain text

Evidence

The inference architecture uses SmolLM2-135M-Instruct in an ONNX-based pipeline, with Transformers.js for model execution and WebGPU acceleration where supported.

Source

The application separates its content workflow from model execution. This lets the workspace remain useful for editing, organizing, scheduling, and exporting drafts even when AI inference cannot run.

Source

The project is built with React, TypeScript, Vite, and Tailwind CSS.

Source

  • Record date: 2026-10-22T00:26:25.754Z