Skip to content

An independent research and product lab based in Barcelona.

AI that brings people and applications closer.

We combine AI research with thoughtful design to make complex tools easier to use and help people work with greater clarity, efficiency and control.

Our first initiative, Living Dossier, is a lawyer-supervised workspace for reviewing residency documents. Currently in development.

Research-to-product loop

Illustrative system map, not a product screenshot

Barcelona lab
Illustrative workflow from documents to reviewed findingsA document panel, evidence links, uncertainty markers and a human review checkpoint.human review

Built Around People

Human control is a product requirement.

01

Human expertise guides the work

People can review, correct and decide. The system is designed to support judgement, not bypass it.

02

Clear interactions

AI helps people navigate complex information and applications through intuitive experiences that make the next useful action visible.

03

Useful outcomes

We evaluate usability, task completion, correction effort and time spent alongside technical performance. Efficiency and productivity are intended outcomes until measured.

From Research To Products

Broad research ambition, grounded in a first concrete initiative.

INtellumina develops AI methods and translates them into practical products for real-world problems. Living Dossier is the first initiative being shaped through this approach.

  1. 1Understand the problem
  2. 2Define the research challenge
  3. 3Build and test
  4. 4Evaluate with practitioners
  5. 5Develop a pilot

Research Foundations

Directions for making AI inspectable in real work.

Our first research focus is document intelligence: connecting findings to their sources and supporting expert review. This work forms part of our broader approach to human-centred AI.

Current development

Understanding complex documents

Working with layouts, multilingual text and relationships across fragmented information so evidence can be organised into something practitioners can inspect.

Current development

Connecting findings to evidence

Keeping source references and visible uncertainty close to generated findings so reviewers can understand why something appears and where it came from.

Planned research

Evaluating AI in real workflows

Building expert-reviewed reference cases and measuring the effort needed to inspect, correct and approve outputs in practical settings.

First Initiative

Living Dossier

In development

A lawyer-supervised workspace being developed for immigration practices to organise residency documents, identify missing or uncertain evidence, and review findings alongside their sources.

Initial scope

  • One residency procedure
  • One matter
  • A source-linked checklist
  • Lawyer correction
  • Reviewed export

Illustrative preview, not an implemented screenshot

Residency evidence checklist

Matter view supervised by a lawyer

uncertainty visible
Proof of residencesource linked
Translated certificateneeds review
Financial evidencemissing item

Review action

Approve, correct or export only after lawyer review.

Trust Is Part Of The Product

Design commitments before launch claims.

These are commitments and planned controls for product development. They are not presented as verified certifications, completed compliance or infrastructure guarantees.

  • Supporting evidence and visible uncertainty
  • Human review, correction and approval
  • Appropriate access permissions and controlled sharing
  • Data minimisation and defined retention/deletion
  • Traceable system actions, sources, corrections and approvals

Partnerships

Collaboration we are ready to discuss.

Immigration practices

  • Domain expertise and workflow evaluation
  • Supervised pilots

Research and infrastructure partners

  • Joint research and experimental development
  • AI infrastructure and deployment

Early-stage investors

  • Financing development milestones

Contact

Discuss a partnership.

Email us about domain expertise and workflow evaluation, joint research, AI infrastructure, supervised pilots or financing development milestones.

Email

hello@intellumina.com

Location

Barcelona, Spain

No home address or private correspondence address is published.