Antwerp · Benelux
Proof of Impact
01 / The mission · DataToday
Buildwise

Technical knowledge that moves from documents to decisions.

Extracting and validating complex technical information for source-linked answers and roof proposals.

Professional ServicesSource-reported delivery
Challenge
Information was spread across text, complex tables and cross-document references.
Our contribution
The ATG route reads documents with Azure Functions, extracts technical values and applies schema control.
Outcome
Minutes — Reported time to a validated roof proposal
Evidence status
Source-reported result · qualitative, sample not specified

Two document-intelligence routes share technical tests and expert validation: structured ATG extraction and source-cited knowledge answers. The source reports roof proposals in minutes rather than months, without a specified measurement sample.

02 / The challenge

What was holding them back?

Information was spread across text, complex tables and cross-document references. Experts needed reusable values and answers whose source they could inspect.

03 / The ambition

What needed to change?

Extract technical values consistently, preserve references and validate output against expert expectations. Reuse the resulting information for technical questions and roof proposals.

04 / Zero Gravity in action

The work, step by step.

  1. DATA

    Azure Functions reads ATG documents into structured extraction.

  2. MEANING

    Values, tables and cross-references become reusable knowledge.

  3. DECISIONS

    Experts test outputs against a fixed reference set.

  4. ACTION

    Validated data supports roof proposals and source-linked answers.

  5. IMPACT

    The reported result and its evidence follow under The impact.

05 / The solution

What we actually built.

Explore the solution

The ATG route reads documents with Azure Functions, extracts technical values and applies schema control. The question-answering route searches text, tables and images and supplies a cited answer. Both routes meet technical tests and Buildwise expert review against a fixed reference set. Re-uploads and overwrites are checked against the agreed structure.

Two routes converge at a common quality-control gate.
  1. 01ATG document → Azure Functions → structured values
  2. 02Schema control
  3. 03Technical question → retrieval across text, tables and images
  4. 04Source-cited answer
  5. 05Technical tests and expert validation
06 / The outcome

What changed.

The source reports validated roof proposals in minutes instead of months. It does not specify an accuracy rate or sample. Structured extraction, schema control and expert evaluation are the documented delivery mechanisms.

Reported resultMinutes

Reported time to a validated roof proposal

qualitative duration; sample not specified

Evidence

DataToday internal delivery portfolio, pages 2 and 11–12.

Baseline: Months

  • Expert reference-set validation and technical tests documented.
07 / Lessons for the next mission

Why it matters.

Document intelligence earns trust through inspectable evidence and reusable structures. Expert validation is part of delivery, not an optional final check.

08 / Technology & capabilities

The enablers.

Explore technology & capabilities

Technologies

  • Azure Functions

Capabilities

  • AI & Machine Learning
  • Data Engineering
09 / Related missions

Related cases.

Your next move

Different challenge.
Same ambition to move forward?

Discuss how the delivery scope relates to your own priorities and operating context.

Start a mission