AI compiler

Compile AI into tiny models.

Describe what your model needs to recognize. Build it through conversation, test it on your images, and refine it in your own workspace.

Compile your project How it works

A segmentation model, built from a description.

01 / THE CONSTRAINT

Foundation models do not solve every deployment problem.

Many AI use cases are constrained by network access, compute, privacy, or data.

TODAY

Train or call

Curate a dataset and train on expensive GPUs, or call a frontier model and pay for every image.

THE COMPILER LOOP

Compile once, run locally

Describe the task and compile a specialist without a new training run. Evaluate it on your own data before you deploy.

02 / THE COMPILER

New tasks. No new training run.

Foundation models already contain rich representations of the world. The ijk compiler turns that learned structure into models for specific tasks.

Define what matters in natural language, then test and refine the result without retraining the underlying model.

Align once

Train a small model to speak the shared representation.

Define in text

Create new tasks without a task-specific labeled dataset.

Compile

Produce a specialist without another training cycle.

03 / YOUR WORKSPACE

Build a model. Keep improving it.

Keep conversations, datasets, and models together in a private project. Save each version, evaluate it on your images, and compare the results.

01DescribeTell the agent what your model needs to recognize.
02CompileBuild and refine the model through conversation.
03EvaluateTest on your images and compare saved versions.
04UseExport supported models to ONNX or serve a saved version through the API.

Compile your project

ONNX export supports classification and segmentation. Detection and segmentation are currently available for non-commercial research evaluation.

VISION TASKS Classification Detection Segmentation

04 / BUILT TO SHIP

The founder has built and shipped edge AI at scale.

PASSIO NUTRITION-AI

Real-time food recognition, entirely on device

Developed edge neural networks that recognized more than 4,000 food types at up to 30 FPS. Used by millions through partners including MyFitnessPal and Elevance Health.

4,000+: food types
30 FPS: on device
Millions: of users

MINDSEYE

Natural language to edge models

Compiled text-defined image classes into downloadable edge models.

~30 MB / 2,000 sign-ups

BYTEGPT

A minified language model

Built a compact language model designed for edge AI.

~85 MB / 1,000 Hugging Face downloads

QUANTUM BASE

Quantum authentication

Customized a contrastive neural network for quantum authentication.

Head of Software, Quantum Base PLC

James Kelly, PhD • Founder, ijk.ai • Former CTO, Passio

Google for Startups Alumni member with access to mentors across Google.

05 / THE THESIS

Foundation models have already learned rich representations from vast amounts of data.

ijk compiles that learned structure into small models for specific tasks.

The resulting models solve real problems in environments constrained by compute, connectivity, privacy, or cost, where running a foundation model is impractical.

06 / GET STARTED

Compile your
first project.

Start with a task you need to solve. Build a model, see how it performs, and make it better.