AI-native studio · 8 products

AI, built into tools you actually use

Desktop-first and local-capable. Typesetting, inference, datasets, ops, workflows — each product solves one concrete problem.

  • Desktop-first
  • Runs locally
  • Your data stays with you
  • Windows / macOS / Linux
  • Open repositories
  • Works offline
  • Deterministic logic in code
8 products
4 categories
7 open repos
4 Platforms
Products

Products

Each one targets a concrete scenario, from content creation to model deployment.

Easy Post

Live

Social content creation, made simple

A desktop content studio with four workbenches — AI ideation, material organization, WeChat article formatting, and social card generation. Typesetting runs on a local Rust engine, not a model.

4 workbenches22+ layout themes1–10 split cards
Windows · macOS Open product

Dr T

Live

An AI ops workbench that actually runs the commands

SSH, AI agents, and accumulated knowledge in one desktop tool. Ask mode suggests and executes in one click; Agent mode loops through tool calls on its own. Works with Ollama and OpenAI-compatible endpoints, so data never has to leave your network.

Multi-host SSHAsk and Agent modesOne-click execution
Windows · macOS · Linux Open product

Easy Infer Station

Live

Cross-platform desktop YOLO inference

Three input sources — images, video files, and RTSP streams — with automatic CUDA acceleration. ROI alarms, fall detection, ByteTrack, and support for any Ultralytics .pt model, ready out of the box.

Images, video, and RTSPYOLOv8 detection and poseROI zone alarms
Windows · macOS · Linux Open product

Easy Dataset

Live

See your dataset clearly, clean it properly

A desktop dataset manager for deep learning. Visual browsing and statistics, annotation quality checks, YOLO / COCO / VOC conversion, and batch label editing — built to work with Easy Trainer Station.

Visual dataset browsingAnnotation quality checksYOLO / COCO / VOC conversion
Windows · macOS · Linux Open product

Easy Conda

Live

Conda environments across Windows and WSL, in one place

A cross-platform Conda manager built with Tauri 2, React, and Rust. Manage environments, packages, and channels across Windows and WSL, with environment.yml import/export, offline delivery, and diagnostics.

Windows and WSL targetsCreate and clone environmentsPackage install and upgrade
Windows · macOS · Linux Open product

Easy Trainer Station

In development

YOLO training without the command line

A desktop workstation for YOLO training (in development). Configure hyperparameters visually, watch Loss and mAP in real time, manage training history and model versions, and work directly with Easy Dataset.

Visual hyperparametersLive Loss / mAP chartsTraining history
Windows · macOS Open product

pt2rknn

Live

PyTorch and ONNX to RKNN in three steps

A conversion tool for Rockchip NPUs with a web UI and no CLI. Supports the full YOLOv8 family (Det / Seg / Pose / OBB), ResNet, and RetinaFace, with INT8 quantization, a calibration set builder, and real-time SSE logs.

Three-step card UIYOLOv8 Det / Seg / Pose / OBBPT to TorchScript to RKNN

GeniusFlow

Beta

Digitize your process, with AI at every step

An enterprise process platform with AI built in. Business users design workflows visually, and each node embeds an AI agent that guides execution, recommends knowledge, and checks quality automatically.

Drag-and-drop designerPer-node AI agentsKanban and Gantt
Who we are

About the studio

Fire Wood Studio is a small team building AI-native products. We believe great AI tools should combine the energy of fire with the grounded patience of wood.

We understand where large models actually help and where they do not, and turn that understanding into tools people can use every day.

Local-first · Deterministic logic stays in code · Your data stays on your machine

Desktop-first
For heavy local work, a desktop app beats a web page: files are read directly, with no upload wait.
Determinism in code
Layout, numbering, and validation have exactly one right answer — so they run in a local engine, not a model.
Data stays local
Local models and self-hosted endpoints keep sensitive data and internal material inside your network.
Get in touch