Late-Night Food Order Orchestration Tool-Use Trajectory Dataset

#tool use #multi-step task execution #order workflow orchestration #late-night food delivery #online food ordering #intelligent ordering assistants

Records: 5000
Size: 0.5GB
Format: JSON
Provider: MOBIUSI INC
Updated: v1.02026-11-05

AI Analysis & Value Prop

This dataset captures end-to-end tool-use workflows for late-night food delivery, from interpreting a user's order request and searching for restaurants open at night and their menus to managing a cart and submitting an order. It records tool calls, returned observations, and subsequent actions, alongside organized information such as user intent, search results, cart state, submission outcome, and workflow status. Use it to train and evaluate models that invoke tools step by step and advance ordering tasks based on tool feedback.

Overview

Dataset Insights

Preview

Sample Examples

sample.json

{}

Schema

Technical Specifications

Field	Type	Description
file_name	string	File name
raw_trace	string	The original trace text, preserving the user's food order request, tool calls, tool observations, and their sequence.
cart_state	string	Cart contents, item quantities, and order totals organized from the raw trace.
next_action	string	The next ordering step inferred or recorded from the final tool observation.
user_intent	string	The user's food requirements, preferences, and delivery instructions extracted from the raw trace.
order_submitted	boolean	Whether the order was successfully submitted, as determined from tool observations.
tool_call_count	integer	The total number of tool calls counted from the raw trace.
workflow_status	string	A summary of the current or final order orchestration status, such as completed, in progress, or failed.
menu_lookup_result	string	Restaurant menu items, available options, and prices organized from the raw trace.
order_submission_result	string	The order submission status, order details, or failure reason returned by the tools.
restaurant_search_result	string	Results for restaurants open at night and related filtering information, organized from the raw trace.

Trust

Compliance Statement

Authorization Type

Proprietary - Commercial AI Training License (No Redistribution)

Commercial Use

Requires exclusive subscription or authorization contract (monthly or per-invocation charging)

Privacy and Anonymization

No PII, no real company names, simulated scenarios follow industry standards

Compliance System

Compliant with China's Data Security Law / EU GDPR / supports enterprise data access logs

FAQ

Frequently Asked Questions

Which ordering steps are covered?: The trajectories cover interpreting user requests, searching for restaurants open at night and their menus, creating or updating a cart, and submitting orders. They also record tool calls, tool observations, and subsequent actions.

What structured information is included?: Alongside the raw tool trace, fields include user intent, restaurant and menu search results, cart state, order submission results, tool-call count, submission status, workflow status, and the next action.

Which model tasks can use this dataset?: It supports tool-use training and evaluation, particularly for learning how to select tools based on an order request, interpret returned information, advance a multi-step workflow, and recognize whether an order is submitted, in progress, or unsuccessful.

How can I access the data and check its license?: Check the dataset listing for available access methods, file specifications, and update information. The publisher's license terms determine permitted uses, including commercial use and redistribution.

Can't find the data you need?

Post a request and let data providers reach out to you.

Get this Dataset

High-quality data, ready for your next AI breakthrough.

Cite this Work

@dataset{Mobiusi2026,
  title={Late-Night Food Order Orchestration Tool-Use Trajectory Dataset},
  author={MOBIUSI INC},
  year={2026},
  url={https://www.mobiusi.com/datasets/188ea96371164ee6a8ea3fd2ec9dd481?dataset_scene_cate_type=2},
  urldate={2026-10-09},
  keywords={late-night food orders, food tool-use trajectories, food ordering training data, late-night restaurant search, menu lookup data, shopping cart state, order submission traces, tool-use training},
  version={1.0}
}

Using this in research? Please cite us.

placeholder

Scene

Domain

Modality

Task