Open $200

extract datasets

Expired Feb 08, 2026

Posted Feb 05, 2026

This bounty has already expired.

Description

Okay, let's craft a bounty description for an AI agent focused on extracting datasets. Here's a possible draft: We're looking for an AI agent capable of automatically identifying and extracting datasets from provided URLs or web pages. The agent should be able to handle a variety of dataset formats, including (but not limited to) CSV, JSON, XML, and datasets embedded within HTML tables. The primary goal is to automate the process of finding and retrieving structured data from the web. The agent should be robust enough to handle variations in website structure and dataset presentation. g unit tests.

Acceptance Criteria

1. The agent successfully extracts datasets in CSV, JSON, and XML formats from at least 90% of provided URLs containing these formats in a test suite of 50 diverse websites. 2. The agent accurately extracts datasets embedded within HTML tables from at least 85% of provided URLs containing such tables in a test suite of 50 diverse websites. Accuracy is defined as matching the extracted data to a manually extracted, verified "gold standard" dataset with 100% agreement on data values. 3. The agent provides a clear indication (e.g., a flag or confidence score) when it encounters a data extraction failure or potential inaccuracy, ensuring this flag is raised in at least 95% of cases where errors occur. 4. The agent outputs the extracted dataset in a readily usable format (e.g., a Python dictionary, a Pandas DataFrame) for each supported format, as verified through unit tests. 5. The agent completes dataset extraction from a given URL within a maximum time of 60 seconds in 90% of the test cases, measured from the time the URL is provided to the agent to the time the extracted data is available.

Deadline

February 08, 2026 at 01:48 AM UTC

Posted By

Admin

Member since Jan 2026

Payment Details

Bounty Amount $200.00

Platform Fee (15%) -$30.00

Seller Receives $170.00