AcuVizionAIAI research & review · Belgium

ACUVIZIONAI / BELGIUM

AI tools for
research and review.

We develop AI models for scientific research and tools that help teams check the evidence behind AI results.

See our work Discuss your project
Early-stage research & development
DATA → MODEL → PREDICTION
Illustrative scientific prediction workflowThree stacked geometric layers represent a structure, followed by a feature vector and a prediction. This is a conceptual diagram, not a measured structure or model output.FEATURESPREDICTIONSTRUCTURE
Illustrative workflow · not a measured structure or result

WHAT WE ARE WORKING ON

Two areas of work.

SCIENTIFIC & MATERIALS AI

Predictions for research.

We develop models that use composition and structure data to estimate material properties. Explore our public materials-research example.

GUARDIAN / EVIDENCE REVIEW

Evidence for review.

We are developing tools to check supported AI results against supplied evidence and declared evaluation rules, with findings for human review.

RESEARCH YOU CAN INSPECT

Public code and results.

View the technical details
Open
research.Follow the work

Reproduce the public example.

Our materials-prediction package includes code and instructions to reproduce the reported outputs.

Submitted for review

The JARVIS submission is open in the active leaderboard repository. Benchmark acceptance is pending.

Code & reproduction Submission status

ABOUT ACUVIZIONAI

Independent research.
A direct conversation.

Founded by Kamran Soleimani in Belgium.

AcuVizionAI is an independent AI initiative led by Kamran Soleimani in Belgium. Our work brings together scientific prediction and evidence review.

Meet the founder on LinkedIn

LET’S WORK ON A REAL QUESTION

Bring your question.
Let’s discuss the next step.

For researchers and teams working with AI results.

Discuss your project

Direct contact with the founder
Kamran Soleimani · Belgium