01First focus
Construction
Project histories connecting tenders, estimates, drawings and specifications with variations, delivery and final outcomes.
Estimate → decision → actual outcome
Licensed business data for AI
Real-world data.
Connected context.
We’re developing a sourcing and licensing route for the records behind real work: the evidence, decisions and outcomes that could help AI teams explore how businesses operate.
CONSTRUCTION FIRST · EXPANDING BY BUYER BRIEF
CASE HISTORY / LINKED VIEWCONCEPT
01
EVIDENCEINPUT
What was known at the time?
Source records · observations · constraints
02
DECISIONCONTEXT
What changed, and why?
Choices · approvals · exceptions · changes
03
OUTCOMEEVIDENCE
What actually happened?
Results · checks · resolution · follow-through
ILLUSTRATIVE STRUCTURE · NOT A LIVE DATASET
LINKED RECORDS & OUTCOMESPROVENANCE & RIGHTS REVIEWREQUIREMENTS-LED SOURCING
01 / Where we’re looking
Starting in construction.
Looking across real work.
Useful context lives across industries. Our focus is a well-documented case of work: the information available, the decision taken and the outcome that followed.
02Exploring
Engineering & manufacturing
Design changes, work orders, specifications and quality records linked to production or validation evidence.
Specification → change → quality check
03Exploring
Maintenance & repair
Service cases linking reported symptoms and incoming tests to diagnosis, intervention and a matched retest.
Fault → diagnosis → repair → retest
04Exploring
Logistics
Manifests, dispatch records and delivery evidence, with the decisions behind shortages, delays and resolved exceptions.
Dispatch → exception → resolution
05Exploring
Commerce & operations
Quotes, orders, inventory and returns connected to approvals, customer-service decisions and documented outcomes.
Request → order → decision → outcome
These are sourcing priorities. Availability, suitability and permissions are assessed source by source.
02 / Our first focus
Construction, connected
from estimate to outcome.
Construction is our starting point. We look for linked project records that show what was expected, why decisions changed and what happened in practice. Explore one illustrative structure below.
SOURCE CONTEXT / ESTIMATEIllustrative
Understand the starting assumptions.
Tender documents, quotations and estimating sheets can capture how a contractor defined scope, allowed for uncertainty and built up a price before work began.
Scope of worksQuantities & ratesAssumptionsRevision history
Potential record types depend on each archive, available rights and agreed preparation.
01
Exploring linked records & evidence
Assess whether source records could support cross-document understanding, reconciliation and reasoning, with original context and available outcomes preserved.
02
Exploring expert-reviewed tasks
Where suitable, assess whether decision-point tasks could be developed, with evidence-backed answers, permitted alternatives and grading criteria agreed with domain experts engaged for each project.
03
Workflow & exception evaluation
Explore whether suitable records could support evaluation of incomplete information, changing requirements and recovery paths. Historical records and executable environments require different scopes.
03 / A considered route
Start with the task.
Build the brief around it.
We’re at the sourcing and buyer-discovery stage. The first conversation is about what you need, so any potential archive can be assessed against a clear purpose.
Define useful data
Your target tasks, sector, record types, formats, coverage and evaluation criteria.
Assess candidate archives
Review the availability, structure, completeness and connections between records and outcomes.
Establish a viable scope
Consider ownership, licensing authority, confidentiality, personal data and restrictions.
Set the next step
If there’s a fit, agree any sample process, preparation, required expert review, permitted use and commercial terms.
Availability and licensing are subject to archive-specific review and agreement. No ready-to-license inventory, exclusivity, legal clearance or model-performance outcome is promised.
04 / Useful to know
A clearer starting point.
For AI labs, applied research teams and buyers exploring specialist business and industrial data.
Do you have a dataset ready to purchase?
Maevar is an early-stage sourcing initiative. Construction is our first focus, with further sectors explored against buyer requirements. We are assessing possible sources; availability, scope and delivery would be confirmed for each opportunity.
What formats or scale can you supply?
That depends on the source and the preparation agreed. Tell us which file types, schemas, volume, date range and case coverage would be useful. We can assess whether linked records or expert-reviewed task sets could be developed, including any expertise and preparation needed, before making a commitment.
How will rights and confidential material be handled?
Each potential source needs its own review of ownership, authority to license, third-party restrictions, confidentiality and personal data. Exclusions, redaction or other controls may be necessary. A potential source is not treated as cleared simply because it is available.
Can we discuss a small evaluation first?
Yes. We can explore whether a bounded evaluation or sample would be appropriate once a suitable source and the relevant permissions have been identified. Any access and permitted use would need to be agreed first.
Are historical decisions treated as correct answers?
No. A recorded decision is evidence of what happened, not proof that it was correct. Any expert-reviewed tasks would distinguish source facts from retrospective annotation, preserve uncertainty and use an agreed review and scoring process.
What should an initial enquiry include?
A short description of your organisation, sector, intended tasks, required records, preferred format and timescale is enough to begin. Please do not attach datasets or files, or include credentials, confidential information or personal records in your first email.
Let’s define the brief
What does useful
data look like to you?
Tell us your sector, the task you’re working on, the context you’re missing and the requirements the data or evaluation tasks would need to meet.
Email your data brief hello@maevar.comNon-confidential summary only. Please do not attach datasets or files, or include credentials or personal records.