Skip to content
AxisXDATA & EVALUATION
中文

EXPERT DATA. VERIFIABLE WORK.

Expert data.
Verifiable by design.

Tasks, reference artifacts and verification evidence across software, agents, knowledge work and science. Explore 24+ data directions, from ready-to-scope products to original tasks built for your team.

Explore current products
MeridianInteractive workspace example
FIN-024Version-aware revenue reviewOpen task
Meridian professional workspace concept
Workspace preview
THE PRODUCT INDEX

Featured data directions 24

Explore specifications & examplesFilter & compare
  1. 01
  2. 02
  3. 03
  4. 04
  5. 05
  6. 06
  1. 07
  2. 08
  3. 09
  4. 10
  5. 11
  6. 12
  1. 13
  2. 14
  3. 15
  4. 16
  5. 17
  6. 18
  1. 19
  2. 20
  3. 21
  4. 22
  5. 23
  6. 24
An expanding portfolio

Start with an existing product or build original data for a new domain, workflow or capability goal.

CUSTOM DATA PROGRAM

Custom data

Scope data around your domain, task format, input modalities and acceptance criteria.

Software & agents · Knowledge work · Scientific reasoning · Custom & longitudinal
DATA TO CAPABILITY

Task & inputsReference & tracesVerification & acceptance

How we deliver

INSIDE THE DATA

Understand the data through a task.

Inspect inputs, references and verification together.

AX / 19

Chartography

Professional chart reasoning

Read specialist charts using domain conventions and turn readings into reliable calculations and judgments.

CURRENT TASK

Read a frequency-response curve at its corner and a specified frequency, explaining the logarithmic axis.

Meridian task example
MERIDIANTask workspace
Interactive example
Data project / Chartographyfrequency_response.svg
Reviewing
frequency_response.svgTASK MATERIAL

一阶归一化频率响应

0.1-22.040.316-16.041-10.043.16-4.04101.96幅值 / dBω / ωc（对数轴）归一化转折点
幅频参考

题目示意数据 / 非实测曲线

Acceptance focus

Check coordinate understanding, curve selection, computation and numerical tolerance separately.

Sources, references and checks stay with the task
Global expert network workspace example
900+ global experts

Expert judgment. Global reach.

Expertise across finance, medicine, law, scientific computing and engineering. Specialists and data teams turn professional problems into tasks, references and independent review.

Explore the global expert network
Axis Edge Research Lab
The AxisX frontier data lab

Axis Edge. Researching what comes next.

A frontier data lab studying long-term memory, agent collaboration, long-horizon work and scientific reasoning. Original scenarios, task environments and verification methods turn model capability gaps into research and training data.

Explore Axis Edge research

EXPERTISE, ENGINEERED INTO DATA

Specialist knowledge. Usable data assets.

AxisX organizes domain experts, engineers and quality teams around model knowledge, reasoning and action. Work spans task definition, inputs, reference artifacts, execution checks and versioned delivery.

About AxisX
0124+

Specialist data directions

0210+

Task scenarios & delivery methods

03102+

University & research connections

046+

Quality review dimensions

01

Expert qualification

Match authors, reviewers and adjudicators through domain experience and trial tasks.

02

Engineering validation

Connect references, environments and verification rules with each task and version.

03

Delivery management

Calibrate scope with samples and manage review, rework and batch acceptance.

Expertise, demonstrated in concrete work.

Inspect delivery structures and methods across specialist domains, from task design to verification evidence.

Expert-authored STEM challenge sets · workbench concept
STEM · Original challengesCASE / 01
DELIVERY MODELOriginal authoring × reference derivation × verifier-led calibration

Expert-authored STEM challenge sets

Original tasks that combine advanced knowledge, multi-step derivation and verifiable answers beyond direct retrieval.

01Mathematics, physics, chemistry, biology and computer science02Problem, reference, rubric and verifier delivered together03Independent solve, model pilot and contamination-risk screen
≈1.2Koriginal tasks5STEM domains6–18%model pass rate
Long-horizon repository engineering evaluation · workbench concept
DeepSWE · FrontierSWECASE / 02
DELIVERY MODELRepository snapshot × long-horizon trace × CI verification

Long-horizon repository engineering evaluation

Repository exploration, multi-file implementation, regression protection and environment verification organized as adjudicable long-horizon work.

01Unfamiliar repository exploration and multi-file implementation02Patch, tests, logs and recovery trace accepted together03Programmatic verifiers plus independent engineering review
310+repository tasks2–14files changed>96%CI reproducibility
Financial reasoning with source evidence · workbench concept
Finance · Risk & researchCASE / 03
DELIVERY MODELFinance specialists × source evidence × independent review

Financial reasoning with source evidence

Research, valuation, risk and compliance judgments delivered with answers, evidence, boundaries and review state.

01Multi-year, multi-source context with citation spans02Answer, evidence, counter-evidence and uncertainty delivered together03High-risk disagreement routed to finance adjudication
≈4.2Kreasoning tasks48–64Kcontext>97%evidence alignment
Explore all 10 case studies

Every step connects to delivery.

Scroll through the structure of a data delivery.

Explore the delivery system
XMeridianDelivery structure example
task_spec.json
source_materials/
source_version.json
Task → reference → verification01 / 03

A defined task. Connected source material.

Goals, input files and source versions enter one record, keeping professional judgment grounded.

XMeridianDelivery structure example
reference_output/
evidence_links.json
review_record.json
Task → reference → verification02 / 03

Turn expert judgment into inspectable work.

Reference artifacts, evidence links and independent review stay connected through each revision.

XMeridianDelivery structure example
acceptance_rules.json
check_results.json
delivery_manifest.json
Task → reference → verification03 / 03

Completed once. Ready to verify again.

Runtime conditions, acceptance rules and check records travel with the batch for technical review and reproduction.

From sample calibration to batch acceptance.

Agree on capability targets, references, runtime and acceptance criteria before production.

  1. 01

    Define the capability boundary

    Specify the knowledge, reasoning or execution boundary—and what a valid task must prove.

  2. 02

    Construct original work

    Expert authors and verification engineers build the task, reference, environment, rubric and validation logic.

  3. 03

    Solve & calibrate independently

    Expert solves, model pilots, environment reruns and difficulty tiers remove ambiguity, leakage and invalid work.

  4. 04

    Run controlled evaluation

    Manage task versions, trajectories, experiment logs, exceptions, rework and state evidence.

  5. 05

    Blind review & expert adjudication

    Independently review tasks, answers, traces and verifiers; resolve disagreement with source or runtime evidence.

  6. 06

    Release versions & feedback

    Deliver data, environments, verifiers and acceptance records, then carry failure patterns into the next release.

Data methodology Quality & verification Security & trust

Professional work inside Meridian

Tasks, source material and expert judgment in a shared workspace.

Financial evidence — Meridian interface concept

Financial evidence

Keep source documents, cited passages and review decisions connected.

Explore Meridian

Meridian workspace examples

LET’S DEFINE YOUR NEXT DATASET

Start with your model’s next objective.

Choose a direction and share the tasks, scale and runtime. We will align samples, deliverables and acceptance criteria.

Or scope a custom program