From the real world to training data01
DATA COLLECTION
Data collection
Turn scene coverage, equipment and capture requirements into a practical plan that connects data sources with project goals.
Explore collectionABOUT sixuandata
About sixuandata
Connecting real sceneswith AI-ready data.
Founded in 2022 in Nanchang, Jiangxi, China, sixuandata turns project needs into clear data guidelines and reviewable deliverables across collection, annotation and automotive scenarios.
From the real world to training data
01Context
Understand where data is used
02Guidelines
Align tasks and quality criteria
03Delivery
Make results usable and reviewable
WHO WE ARE
Focused on the data work. Mindful of what comes next.
Established in 2022, Nanchang Sixuan Technology Co., Ltd. operates under the sixuandata brand. We provide data collection, annotation and automotive data services for AI research, algorithm and data teams, helping clients define coverage, task guidelines and delivery requirements around their use cases.
We focus on the details that make data usable: aligning requirements through pilot samples, keeping quality review and correction records, and organizing each batch with clear version and delivery notes. This makes the results easier to review, use and improve.
OUR FOCUS
Three service areas. One shared purpose.
From sourcing suitable data to consistent annotation and scenario-specific deliverables. Explore each service and its related cases to understand the methods and outputs.
02
DATA ANNOTATION
Data annotation
Define objects, boundary rules and exception handling so annotation work and quality reviews use a shared standard.
Explore annotation03
AUTOMOTIVE DATA
Automotive data solutions
Connect collection, annotation and delivery around road scenarios and multimodal data, from the initial brief to a structured dataset.
Explore automotiveHOW WE THINK
Principles that shape the work.
Clear agreements, reviewable results and timely feedback form the basis of this working approach.
- 01
Align before execution
Start with the use case, define objects, coverage and acceptance criteria, then use pilot samples to resolve differences in interpretation.
Practical outputRequirements and sample confirmation - 02
Explain quality with records
Translate quality requirements into checks, and keep issue and revision records so decisions can be reviewed later.
Practical outputReview records and issue lists - 03
Bring feedback into the next batch
Turn usage feedback into specific rules, distinguish current corrections from future guideline changes, and keep versions consistent.
Practical outputChange records and version notes
WORKING TOGETHER
A shared project brief connects every role.
Project coordination, collection, annotation, quality review and delivery work from shared guidelines, with responsibilities and communication agreed for each project.
SHARED REFERENCE
Project goals · Data guidelines · Acceptance criteria
Keep every role aligned on the same version, from kickoff to delivery.
Project coordination
Align scope, milestones and communication; organize requirement changes and open questions.
Project plan / decision logCollection & annotation
Carry out data tasks against the guidelines and flag edge cases or samples needing clarification.
Task outputs / issue feedback
Quality review
Review samples and batches against agreed criteria, then track corrections and recheck results.
Review records / recheck resultsData delivery
Check files, formats and manifests, and prepare version notes and delivery scope.
Data files / delivery notesProject working cadence
- 01Align brief and samples
- 02Execute and review
- 03Resolve and confirm
- 04Deliver and collect feedback
LET’S BUILD WITH DATA
Start with a concrete data need.
Share your use case, data type or current challenge, and outline the questions to resolve next.
Talk with usStart with a brief, then align scope, guidelines and delivery.