انتقل إلى المحتوى

HCT Sovereign Agentic AI Platform

Sovereign Agentic AI for Government and Enterprise

HCT Group builds, deploys and governs on-premises Agentic AI for Police, Government and Enterprise across Yemen. Bilingual Arabic and English, full data residency, and hundreds of production-ready AI agents delivered end to end.

  • On-premises deployment
  • Arabic and English
  • In-country data residency
  • Human-in-the-loop governance
Request a DemoExplore the Platform
300+
Agentic use cases delivered per program year
100%
On-premises, no external cloud dependency
2
Languages built in, Arabic and English
24x7
Supported platform with defined SLAs

A governed AI factory, not a chatbot

What Agentic AI means for a serious organisation

Agentic AI is software that does not just answer questions. It plans, uses tools, calls systems, follows approvals, and completes real work under human control. HCT delivers this as a governed platform where every agent has an owner, a risk classification, an approval path, allowed tools, test evidence, a version record, and a full audit trail.

  • Prompt-to-agent creation with a no-code and low-code builder
  • Multi-agent orchestration with routing, retries and state
  • Retrieval over your own documents and systems with citations
  • Connectors to APIs, databases, RPA and enterprise applications
  • Local LLM integration so models run inside your environment
  • Guardrails, human approvals, observability and exportable audit logs

The platform

One platform, eight governed building blocks

The HCT Sovereign Agentic AI Platform is an on-premises enterprise stack. Each layer is designed for public-safety and regulated environments, from the builder your teams use to the audit trail your risk officers rely on.

Agent Studio

No-code and low-code build experience. Prompt-to-agent, templates, Arabic and English interaction, reusable skills and a controlled publishing workflow.

Multi-Agent Orchestrator

Coordinates agents, workflows, tools and approvals with sequential and parallel tasks, task routing, retry logic and state management.

Tool and Connector Gateway

Secure integration layer for APIs, databases, files, email queues and RPA tools such as UiPath where applicable.

Local LLM Gateway

Model abstraction and local model connectivity with routing, prompt policies, token budgets, evaluation and model audit.

RAG and Knowledge Layer

Grounded responses and document intelligence. Ingestion, chunking, indexing, embeddings, retrieval, citations and metadata security.

Governance and Guardrails

Controls AI risk and behaviour with policy controls, Arabic-aware guardrails, safe tool use, approval workflows and department scope.

Observability and Audit

Operational control through run traces, logs, metrics, failure analysis, usage, costs, alerts, dashboards and audit reports.

Lifecycle and DevSecOps

Enterprise operations with versioning, promotion from Dev to UAT to Production, backup and restore, rollback and release control.

Sovereign by design

Your data stays in your environment

For Police and Government, sovereignty is not a feature, it is the starting point. HCT deploys the platform inside your approved infrastructure. No operational data is sent to external cloud services or used to train external models.

Request the security overview
  • On-premises deployment inside your approved zones
  • RBAC with SSO and Active Directory integration
  • Encryption in transit and at rest, secrets management, least privilege
  • Human-in-the-loop approvals with maker and checker controls
  • Arabic-aware guardrails and restricted tool use
  • Full execution trace and exportable audit logs
  • High availability, backup, restore and disaster recovery

End to end for Yemeni organisations

A controlled use-case factory, not a proof of concept that stalls

HCT runs a disciplined delivery model that turns a backlog into accepted, documented, production agents. Every use case moves through the same governed pipeline, managed from a single prioritised backlog you approve.

  1. 01

    Discover

    Workshops, process review, data readiness and risk classification.

  2. 02

    Prioritise

    Rank by value, feasibility, urgency, data readiness and risk.

  3. 03

    Design

    Agent roles, workflow, prompts, tools, integrations and guardrails.

  4. 04

    Build

    Configure agents, connectors, RAG index, approvals and monitoring.

  5. 05

    Test and UAT

    Functional, integration, Arabic, security and user acceptance.

  6. 06

    Deploy and Handover

    Production release, training, runbook and source-code handover.

Knowledge and RAG pod

Search and document agents

Grounded answers, summarisation and knowledge retrieval across your document estate with access-controlled indexes.

Workflow and RPA pod

Process automation agents

Multi-step workflows across APIs, databases and RPA with approval paths and connector evidence.

Complex multi-agent pod

High-control agents

Multi-agent orchestration for sensitive actions with advanced guardrails and security review.

Who we build for

Built for the organisations that carry real responsibility

From public safety to regulated enterprise, HCT maps Agentic AI to the work that matters, with the controls each sector demands.

🚔 Police and Public Safety

Briefing and incident summaries, shift handover, investigations and case support, emergency SOP assistants, bilingual controlled public response, and evidence checklist support with full auditability.

🏛️ Government

Service-desk triage, permit and process automation, policy question answering, tender and contract assistants, and natural-language data queries over departmental systems.

🏢 Enterprise

Finance, procurement and admin automation, contract clause search, HR onboarding and policy agents, and IT service management with human approvals on high-impact actions.

📡 Telecom

Network operations knowledge agents, ticket triage, field report drafting and vendor query handling, aligned with HCT deep telecom engineering heritage.

🏦 Finance and Regulated

Controlled document intelligence, compliance checklists, audit-evidence retrieval and reporting support, all inside your own governed environment.

⚡ Energy and Infrastructure

Operational knowledge search, maintenance and SOP assistants, and reporting agents for asset-heavy operations across Yemen.

Why HCT

End to end, and fully handed over to you

HCT Group Yemen is backed by a regional technology group with the engineering depth to deploy, integrate and govern AI inside demanding environments, and the discipline to hand it all back to your team.

Delivery across Yemen

A resident on-site team supported by an expert pool, covering the full data and AI lifecycle for organisations across Yemen, from Sana'a and Aden to Hadramout.

Full-stack integration

APIs, RPA, databases, local LLMs and legacy systems connected through a reusable connector library, with every integration documented.

Complete handover

Agents, prompts, workflows, configurations, connectors, documentation, training and source code for custom components are handed over to you.

21 mandatory capabilities

On-prem, bilingual, prompt-to-agent, RPA, local LLM, RAG, approvals, guardrails, observability, RBAC and SSO, all confirmed as standard.

Adoption built in

An AI Champion model with bilingual training for builders, administrators, security and support so capability stays with your people.

Governed partnership

Clear milestones, steering governance and defined SLAs, from a group already trusted by leading public and enterprise organisations in the region.
Answers

Agentic AI, answered clearly

The questions Police, Government and Enterprise leaders ask us most.

What is Agentic AI?

Agentic AI is AI that can plan and carry out multi-step work, not just answer questions. Agents use tools, call systems, retrieve documents, and follow approval steps to complete real tasks. HCT delivers this as a governed platform where every agent has an owner, allowed tools, guardrails and a full audit trail.

Is the HCT platform deployed on-premises?

Yes. The HCT Sovereign Agentic AI Platform is deployed inside your approved infrastructure. Your operational data stays in your environment and is never sent to external cloud services or used to train external models.

Does it support Arabic as well as English?

Yes. The platform is bilingual by design, with Arabic and English interaction, Arabic-aware guardrails, and Arabic and English training materials for your teams.

How does HCT keep AI safe and controlled?

Through role-based access, SSO and Active Directory integration, human-in-the-loop approvals, Arabic-aware guardrails, encryption, and full execution traces with exportable audit logs. Each agent has a risk classification and an approval path before it can act.

Can it connect to our existing systems?

Yes. A secure connector gateway integrates APIs, databases, files, RPA tools such as UiPath, and local LLM endpoints. Custom functions are built where needed, with source code for custom components handed over to you.

How quickly can HCT deliver use cases?

HCT runs a controlled use-case factory with specialised delivery pods. A mature program delivers hundreds of accepted, documented and production agents within a single program year, managed from one prioritised backlog you approve.

Do you serve organisations across Yemen?

Yes. HCT delivers end-to-end data and Agentic AI programs for Police, Government and Enterprise across Yemen, in Sana'a, Aden, Taiz, Hodeidah, Mukalla and beyond, from deployment and integration through to training and full handover.

Bring governed Agentic AI into your organisation

Book a briefing with the HCT Agentic AI team. We will walk through the platform, the governance model, and a delivery plan mapped to your priority use cases.

Request a BriefingTalk to Sales