Preliminary Round Results

Alibaba Cloud×Atlas

Agentic AI Hackathon

TOP 10Finalists

Ten finalist teams built autonomous agents on a live travel-booking stack — from group-chat trip coordination and care-plan-driven replanning to multi-agent disruption recovery and true all-in fare verification.

Kick-off Workshop19 AUG 2026

Submission30 AUG 2026

Final RankingTO BE ANNOUNCED

Alibaba Cloud × Atlas Agentic AI Hackathon key visual

Organised by

Evaluation Partner

Listed alphabetically — not a ranking Final ranking to be announced

Project

achievego.ai

Evaluation

Rubric-driven human evaluation of agent output.

Email Password

Enter your account email and we’ll send a password reset link shortly. Please check your inbox and spam folder.

Email

Choose a new password for your account.

New password Confirm new password

This account must set a new password before continuing.

Current password New password Confirm new password

/ Evaluation

—

— —

Workbench

New batch

Pin a published rubric version and a review layout, then choose the samples and the reviewers. Everyone on the batch reviews every sample in it.

New rubric

Questions describe what to observe; the answers form the review.

Import rubric from JSON

Paste a complete rubric object. It is validated the same way the editor is, and saved as a draft version.

Download the schema guide

Rubric JSON

Add to

Compare rubric versions

What changed between two versions of the same rubric.

New account

The generated password is shown once. The owner must replace it on first login.

Name Email Role

Confirm