Development

Ace Developer

Ace AI

Quickstart

Comprehensive guide to the Ace AI API endpoints

Ace API

Authenticates a user by username and password. Returns a session token in an HttpOnly cookie.

Auth Login

Creates a new user account. Requires username, password, role, org, github, and linkedin.

Auth: Signup

Updates a user's profile with expertise, bio, education, and experiences.

Auth: Update Profile

Returns a signed URL that allows authorized clients to upload a dataset to a secure storage bucket.

Synthetic: Upload URL

Returns a signed URL that allows authorized clients to download a previously uploaded or generated synthetic dataset.

Synthetic: Download URL

Creates a request for generating a new synthetic dataset based on an existing dataset. Typically triggers a background process or an agent to generate data.

Synthetic: Generate

Returns a signed URL that allows authorized clients to upload data for Generative AI tasks.

GenAI: Upload URL

Returns a signed URL that allows authorized clients to download data used or generated by Generative AI tasks.

GenAI: Download URL

Provides an AI-based analysis, evaluation, or inference on the uploaded dataset by sending a custom prompt. May be used to run model inference or other tasks on the dataset.

GenAI: Perform

Fetches a random task based on the provided query parameter 'task-type'. Currently supports 'ellipsis-tier2-task'.

Editor: GET /tasks/random

Creates a new task of the specified type. Currently supports 'ellipsis-tier2-task'.

Editor: POST /tasks/{taskType}

Updates fields of an existing task identified by its task type and id.

Editor: PATCH /tasks/{taskType}/{id}

Fetch evaluation data filtered by username, status, limit, offset, and expertise.

Evaluator: GET /evaluation-data

Evaluator: POST /evaluation-data

Returns a random evaluation data entry based on the evaluator's username.

Evaluator: GET /evaluation-data/random

Updates evaluation data fields such as tags, status, data, and rating.

Evaluator: PATCH /evaluation-data/{id}

Fetches comments for a given evaluation data entry using the query parameter evalDataId.

Evaluator: GET /evaluation-data/comment

Creates a new comment for an evaluation data entry.

Evaluator: POST /evaluation-data/comment

Updates a comment associated with an evaluation data entry. The commenter in the request body must match the original commenter.

Changelog