Curiosity, put to work.

Software, hardware, and applied research by Softworks. From Egypt, for problems worth solving.

Explore our questionsAI generated close study of a circuit board.
AI studies & licensed footage

Begin with the place.

The language we speak. The devices we have. The work we repeat every day. Ordinary questions worth time and an experiment.

AI generated fields and irrigation channels beside the desert.
AI studies & licensed footage

Learn something. Leave something useful.

A method someone can repeat. A tool they can adapt. A risk they can understand. That is what we want to share.

From real work to better questionsAI generated young palm growing beside water.
AI studies & licensed footage

Good technology should understand where it lives.

Our language. Our everyday work. The details that get overlooked. We started a lab to give them time.

Why our lab exists

From real work, to better questions.

Our work with collaborators on Transito and DocuBex raises questions about what connects a screen, a device, and a record.

From a scan to a record we can trust.

What happens if a connection fails during a handover? We care about the small steps that let software and hardware work together.

Questions of reliability
An illustrative workflow. No live device or real shipment.

A fictional handheld scanner captures record DEMO-042. When the connection is unavailable, the record stays on the device. Reconnecting sends that same record and returns one handover receipt. This animation is a design example, not a tested product claim.

Work that informs our questions. AI artwork and licensed nature footage.

Problems we want to understand in Egypt.

The OECD’s 2025 review describes uneven use of digital tools among businesses. We want to understand that gap in everyday work. Read the review

35.6%

Reported electrical outages.

Among firms covered by the World Bank’s Egypt 2025 survey, during each establishment’s last complete fiscal year.

Formal private establishments with 5+ employees in covered sectors.

See the World Bank profile
Scope and source

Fieldwork covered 1,024 establishments from September 2025 to January 2026. Coverage is formal private establishments with at least five employees in manufacturing and selected services. It excludes informal and smaller microbusinesses. This is not an AI adoption measure.

Survey data and methodology

The same information, entered again.

Paper, Excel, messages, and disconnected tools. We want to find where information is repeated and what could be connected or automated.

A system that is hard to change safely.

Unsupported software, fragile integrations, and unclear permissions. How can we improve them without stopping the work?

Work that needs to survive an interruption.

Offline capture, retries without duplicates, recoverable backups, and devices we can repair.

Questions to start with

These are proposed lab questions. The figures are external evidence, not findings from our own study.

The result is only part of it.

Findings and risks. Datasets, open projects, and practical kits. Shared with the method and the limits of what we know.

An illustrative release package. Nothing published yet.

An imagined research package brings a manuscript, method, code, and data notes together for review. This demonstration does not represent an existing paper or dataset.

We use AI to research and code. We review the work ourselves.

How we work
AI image study of Western Desert formations.

A place for what we find.

Papers, datasets, and open tools. Our first release is still ahead of us. When it is ready, the work behind it will be here too.

Explore the library