MubtakaratMUBTAKARAT
OperationalLAMH · a global tool: a trust layer for every source your AI relies on ↗

Sovereign AIfor nations and institutions that cannot afford to be wrong.

Mubtakarat builds AI platforms and agents that run where you decide and under your authority: your data is processed in the environment you choose, every answer has a source, and every consequential decision passes a responsible person and is recorded for audit.

01Deploy anywhere

Your premises, your cloud, your region, or fully air-gapped from the internet.

02Human authority

No consequential action without explicit approval from a responsible person.

03Audit by design

Every action logged with its reason, actor and time, and open to review.

04Global by default

Multilingual and multi-jurisdiction from day one.

The systems that run nations and markets cannot afford AI that guesses. So we build technology that proves every answer, respects every boundary, and stands up to any audit.

Mubtakarat for Information Technology
76.0072.00θ 81.6°ABDEM22 · 17+5=22
DWG NO.MBT-M22-001
SUBJECTBRAND MARK · 22 STARS
REFMESSIER 22
01Platforms

Platforms for work that cannot fail.

Each platform solves a specific operational problem and is built to the same standards: data sovereignty, source attribution, human approval and auditability.

Operational

LAMHSOURCE TRUST FOR AI · GLOBAL

Your AI answers with confidence. Make sure its source is still valid.

LAMH certifies every piece of knowledge your AI relies on: intact, current and still valid, before it reaches the user.

  • A validity certificate for every source before it is used
  • Flags and blocks expired or altered sources
  • Works with your existing AI systems, no rebuild
  • Hosted cloud service: data encrypted on servers in the EU (Frankfurt)
Live verificationIllustrative example
Policy handbook 2026Intact · current · in force
Valid
Old price listSuperseded by a newer version
Replaced
File changed without approvalDoes not match the original
Blocked
Procedures circular 14Intact · in force
Valid
Sources checked: 4lamh.mubtkrat.ai
Research & development

The next pipeline

Projects in development. We publish outcomes, not details.

R&D · 01Amin 7

An independent standard for AI-agent reliability, giving organisations provable confidence before they depend on them.

R&D · 02Jathr

A language-aware tokenizer that cuts AI compute and energy costs, especially for the world’s morphologically rich languages. Early results: around 40% shorter inputs for the model.

R&D · 03Bayan

A new-generation programming language that brings code closer to how people think, and leaves less room for error.

R&D · 04Truvo

Financial infrastructure that runs behind digital wallets and banks, making money move faster and at lower cost.

R&D · 05Energy Efficiency

Solutions that cut data-centre energy costs without compromising performance.

R&D · 06Eternal OS

One secure platform that runs an organisation’s everyday financial operations.

02Sectors

Where precision is a matter of sovereignty.

We build for organisations that handle sensitive data and high-consequence decisions, and cannot let either leave their control.

01

Government & public sector

Process automation, document digitisation and service agents, with data kept in-house.

02

Justice & legal

Legislative research that cites the text, case management and document authenticity.

03

Financial services

Compliance, transaction monitoring and faster, lower-cost payment infrastructure.

04

Energy & critical infrastructure

Lower data-centre energy costs, and systems that run in isolated environments.

05

Education & national archives

Offline reading and indexing of records and manuscripts, with secure digital registries.

06

Compliance & risk

Oversight of sources and decisions, with evidence ready for auditors and regulators.

03Deployment & sovereignty

Runs where you decide.

Deployment is a sovereign decision. We fit every platform to your security and regulatory requirements, not the other way round.

  1. 01

    On-premise & air-gapped

    Runs on your servers, can operate with no internet connection at all, and installs from removable media.

    Maximum isolation
  2. 02

    Cloud in your chosen region

    Hosted in data centres in the country or region you choose, with full control over access and keys.

    Data residency where you need it
  3. 03

    Hybrid

    Sensitive data on-premise, general workloads in the cloud, under one security policy.

    Flexibility without compromise
Encryption in transit and at rest Least-privilege access Tamper-evident audit logs Security review before every release Retention and destruction policies
Guided by international frameworksISO/IEC 27001GDPREU AI ActNIST AI RMFNCA ECC

Alignment and preparation by design, not certifications.

04Agent governance

Machine speed. Human authority.

We apply AI agents to our own internal operations, under strict rules that stop any error from reaching a client or money. What we offer our clients, we tried on ourselves first.

01

One task per agent

And a written playbook that sets what it may and may not do.

02

Double review

Every proposal, payment or external message is checked by an independent agent first.

03

Authority limits

Anything above the limit runs only with explicit approval from a responsible person.

04

Full audit trail

Every decision is logged with its reason and who made it, and can be reviewed at any time.

05Capabilities

Integrated engineering capability, under one roof.

Capabilities that cover the whole solution cycle, from infrastructure to operations, combining human expertise and AI agents.

AI & agents

  • Operational agents with authority limits
  • Source-grounded knowledge retrieval
  • Multilingual language models

Platform engineering

  • Government and enterprise systems
  • Integration with existing systems
  • Web and mobile applications

Data & enterprise intelligence

  • Document and archive digitisation
  • Decision dashboards
  • Data governance

Security & infrastructure

  • On-premise and cloud deployment
  • Vulnerability assessment and protection
  • Data-centre energy efficiency
06Engagement model

From briefing to full operation.

A clear path with measurable outputs at every stage. No broad commitment before the pilot proves its value.

  1. Briefing & assessment

    A confidential session to understand the mission, data and constraints, and set success measures before any promise.

  2. Pilot on your data

    A time-boxed pilot inside your environment, measured against the baseline.

  3. Sovereign deployment

    Deployment in the approved environment, with integration, training and knowledge transfer to your teams.

  4. Operate & assure

    Managed operations, continuous updates and an agreed service level.

Institutional pilot programme

A limited number of institutions pilot our platforms on their real data.

We select specific organisations to pilot our platforms before general release, and shape the solution around their mission and requirements.

Apply to join
  • A pilot on your data, in your environmentWe run the platform on your real documents or processes, in the environment you approve.
  • A dedicated teamDirect follow-up from our engineering team throughout the pilot.
  • A say in the roadmapYour requirements set the next development priorities.
  • Measurable resultsWe agree success measures before we start and measure them together.
07Company

Mubtakarat for Information Technology

A technology company building world-class AI platforms for governments and enterprises around the world. Headquartered in Riyadh, serving clients everywhere.

We believe AI will reshape how institutions are run, so we started with ourselves: we apply AI agents to our own internal operations under strict governance, and what we offer our clients we tried on ourselves first.

We build for government, judicial, financial and education bodies and large enterprises, wherever they are.

Company detailsHQ · GLOBAL
Legal name
Mubtakarat for Information Technology Co.
Headquarters
Riyadh
Registration
Legal information
Scope
Global
Email
info@mubtkrat.com
08FAQ

Questions from procurement and security teams.

Q.01Can your platforms run with no internet connection at all?

Yes. Platforms such as Ijam are designed to run entirely inside the organisation’s network and can be installed in fully air-gapped environments, so documents and data never leave the premises.

Q.02Where is our data stored, and who can access it?

Wherever you decide: on your premises or in a cloud in the region you choose. Access follows least privilege, and every access is logged for audit. The hosted LAMH cloud service keeps data encrypted on servers in the EU (Frankfurt).

Q.03How do you stop AI from making answers up?

Answers are grounded in defined sources that are cited with every answer, and when there is no source the system says so plainly. Consequential decisions go to a responsible person before they are carried out.

Q.04Do you sign NDAs and work within procurement processes?

Yes. On request we sign a non-disclosure agreement before any detailed briefing, and we work within the organisation’s procurement and contracting requirements.

Q.05What is the support model after deployment?

Technical support, managed operations and continuous updates, at a service level agreed with each organisation.

Q.06Where are you based, and who do you work with?

Mubtakarat is headquartered in Riyadh and serves public and private organisations anywhere in the world, remotely or on site depending on the mission.

09Contact

Request an executive briefing.

Tell us about your organisation and the mission, and our team will arrange a confidential briefing with the right specialists.

What happens next
  1. 01
    We review your requestReviewed directly by the engineering team.
  2. 02
    We arrange a briefingA focused session on your mission and requirements, under NDA on request.
  3. 03
    We propose a pilot scopeWith clear success measures before any commitment.
Emailinfo@mubtkrat.com