Real-world knowledge.More capable AI.

Find data that reflects how work actually gets done. We connect a precise model or agent requirement with suitable sources and an agreed path to licensing.

Find my data

For AI labs, applied AI companies, research teams and builders of domain-specific agents.

Example output

Expert knowledge for AI

Situation: Invoice price differs from the order
Relevant facts: Order: 120 units × USD 18.00. Invoice: 120 × USD 19.50. No approved price change.
Expert procedure: Compare the order and invoice; check amendments; flag the USD 180.00 difference; request a corrected invoice.
Resolution: Corrected invoice matches the approved order, or an authorized amendment explains the difference.

The useful data is the reasoning and sequence of actions, not just a document title or a generic answer.

Clearly defined fieldsCSV / JSON

Data for AI

What you receive

Task examples with context, expert actions and expected outcomes
Source references, language and domain for each record
A structured delivery for retrieval, training or evaluation
Agreed usage rights and documented preparation steps

A focused first step

Start with one business task, a defined domain and your chosen languages. We assess source rights, useful examples and the format your AI team needs.

How the engagement works

Individually scoped sourcing, preparation and licensing. Provider remuneration and any brokerage fee are agreed transparently.

Find my data

WHAT HABITABLE DELIVERS

Less explaining. More actual data.

See the fields, records and business question behind this example.

Example output

Expert knowledge for AI

CSV / JSON

The business question

“How should an AI assistant resolve this situation, step by step?”

What we collect

The task, relevant facts, expert checks, the recommended action and the resolution criteria — prepared from appropriately licensed source material.

What it helps you do

Give AI teams structured examples for retrieval, evaluation and training around a specific business task.

Scroll horizontally to see all fields

Expert knowledge for AI — Example output
Situation	Relevant facts	Expert procedure	Resolution
Invoice price differs from the order	Order: 120 units × USD 18.00. Invoice: 120 × USD 19.50. No approved price change.	Compare the order and invoice; check amendments; flag the USD 180.00 difference; request a corrected invoice.	Corrected invoice matches the approved order, or an authorized amendment explains the difference.
Customer needs a new delivery address	Order is paid. The warehouse has not released the shipment.	Verify the customer; check dispatch status; update the address before release; confirm the change.	The shipment uses the confirmed address and the customer receives a confirmation.
Two supplier records may be duplicates	Names differ, but the company registration number and website domain match.	Compare legal identifiers and addresses; check branch relationships; merge only after confirming the same entity.	One supplier identity retains both original record IDs and a documented merge decision.

The useful data is the reasoning and sequence of actions, not just a document title or a generic answer.

Request similar data Download CSV

Example records showing the structure of a delivery. Sources, markets and scope are agreed for each project.

FROM A DATA GAP TO A WORKING DATASET

Bring the question.We’ll connect the dots.

Define the outcome

Tell us what decision or product the data should support. We agree what useful looks like.

Find the right sources

We assess collection, existing datasets and direct partnerships, including access and usage rights.

Validate a sample

Review the structure, quality and limitations before committing to a larger delivery.

Deliver. Keep it useful.

Receive data in the agreed format, with refreshes and quality checks where you need them.

LET’S CONNECT THE DOTS

Your next dataset starts here.

A question, a missing field, a new capability.Tell us what you want to build.

Find my data Explore data partnerships