Casablanca and the Hassan II Mosque at sunset, viewed from the coastline.

Enterprise AI consulting & engineering

Deploy. Govern. Scale.

Hunter BI helps business leaders, CIOs and AI teams connect assistants to their operating systems. Strategy, engineering, governance and adoption — from a scoped departmental pilot to an evidence-led rollout decision.

Explore the corporate pilot

Credentials & badges

Expertise you can verify.

Architecture, deployment and adoption: OpenAI and Anthropic learning paths supporting your AI projects.

Earned byHamza El Hinanifounder of Hunter BI

Credentials are awarded to an individual. Visit the issuer’s record to check the recipient, scope and validity.

Explore all 11 credentials and badges

Selected client integrations

Enterprise systems. Connected to AI.

Completed Hunter BI integrations across ERP and marketing operations. Each summary separates confirmed delivery from the results and technical detail available for publication.

Completed client integration

SAP

ERP / business operations

MCP integration enabling a client to work from an AI assistant within the project's SAP scope.

Read the integration brief

Completed client integration

Sage X3

ERP / business operations

Client support connecting Claude AI to Sage X3 through a custom-developed MCP integration.

Read the integration brief

Completed client integration

Odoo

ERP / connected workflows

AI/MCP integration for Odoo completed for a client, as confirmed by Hunter BI.

Read the integration brief

Completed client integration

Meta Ads Manager

Marketing / Facebook advertising

AI/MCP integration for Meta Ads Manager (Facebook Ads Manager) completed for a client, as confirmed by Hunter BI.

Read the integration brief

Completion is reported by Hunter BI. Client identities, project documents and independently verified performance figures are not published. Software names identify the connected tools, not an endorsement by their publishers.

Your business environment

The tools your teams already use.

SAP, Sage X3, Odoo and Meta Ads Manager are represented in our completed integrations. For the rest of your environment, we assess each application, its interfaces and the permissions needed for your workflow.

Explore the possibilities

One request. Your tools in motion.

Choose a workspace in the scene.

Six business workspaces connected to a Hunter BI AI assistant: CRM, email, knowledge, projects, data and marketing.
01 / 06

Customer relationships

Integration to assess

You ask

« Prepare my client meeting using the opportunities and recent conversations I am authorised to access. »
  1. 1Read the CRM
  2. 2Review conversations
  3. 3Prepare the brief

The assistant would prepare

A structured sales brief covering context, open opportunities and discussion points, with references to the information consulted.

Control to configure : Access restricted to the agreed sales scope. Any CRM update requires approval.

Scope this workflow with Hunter BI

Illustrative scenarios, with no client data or live connection. Feasibility and permissions must be assessed for your environment.

Our completed integrationsSAP Sage X3 Odoo Meta Ads Manager
View the tools and integration prerequisites

Integration to assess

Customer relationships

Salesforce · HubSpot · your CRM

Qualify customer-data access and a sales workflow.

Integration to assess

Email & productivity

Microsoft 365 · Google Workspace

Define the authorised mailboxes, calendars and documents.

Integration to assess

Knowledge & collaboration

SharePoint · Notion · Teams · Slack

Identify approved sources and user access boundaries.

Integration to assess

Projects & operations

ClickUp · Jira · your internal applications

Scope task preparation and the updates that require approval.

Integration to assess

Data & reporting

Power BI · Excel · your databases

Agree the indicators, reference data and access rights.

Integration to assess

Marketing & acquisition

Google Ads · LinkedIn Campaign Manager · your marketing stack

Separate reporting from changes that affect publication or spending.

Not every tool requires a custom MCP server. We assess an existing connector, an authorised API or another supported interface first. Availability depends on the product, version, licence and permissions; there is no universal connector promise.

Assess your business tools

What we deploy

Two flagship platforms. One governed method.

ChatGPT Enterprise

Scope the workspace, access controls, included connectors and adoption plan against your licence and business requirements.

Claude Enterprise

Anthropic’s enterprise offering for document work, governed team adoption and integrations, with capabilities checked against your use cases.

Agents & automation

Business agents with defined tools, source verification and approval gates for consequential actions.

Governance & sovereignty

Data-flow mapping, access controls, deployment choices and evidence for your security and compliance review.

Explore the detailed platform and service pages in English, or contact us to discuss your organisation's context.

Official partner networks

OpenAI Partner NetworkAnthropic Partner Network

Enterprise AI consulting in Morocco

Hunter BI is an AI consulting and engineering company based in Casablanca. We help organisations choose useful AI applications, connect assistants to business software and establish the controls needed to operate them. Our work brings together deployment strategy, integration, governance and team training.

We start with a task and the people responsible for its outcome: finding an answer in internal documents, preparing a business record or assisting a developer. The first decision is whether AI is useful for that task at all. The agreed scope defines the data, access rights, acceptance tests and human approvals before implementation.

Your software. Your AI assistant.

A conversation becomes useful when it can reach the right information through authorised interfaces. We develop MCP servers and API integrations for business systems, keeping the existing application as the source of truth. The assistant receives a defined set of tools, not unrestricted access to an ERP, a database or a workstation.

MCP does not create an interface that the software does not provide. We inspect the installed version, available services, licensing and permissions, then test the selected assistant and connection method. Reading information, preparing a change and executing it are separate operations with separate controls.

ERP, CAD and 3D: work from the assistant

Hunter BI has completed MCP integration projects involving SAP, Sage X3, 3ds Max and SOLIDWORKS. The reported benefit is that the client can work from an AI assistant with improved productivity. This is qualitative feedback: no independently measured percentage, financial return or named client testimonial is published.

The dedicated pages distinguish that experience from proposed use cases for a new project. Your software version, documents, business rules and acceptance criteria determine what can be delivered. A successful connection in one environment does not establish universal compatibility.

Your data, your access rules

The deployment decision covers more than the model. We map what information is sent to the assistant provider, which accounts can access each source and how outputs are retained. Hosting a connector inside your network does not automatically keep the model's processing there.

For sensitive workloads, we assess private infrastructure and open-model options alongside cloud services. Governance defines owners, permitted uses, escalation rules and evidence to retain. A technical configuration supports a compliance review; it does not replace the judgement of your legal and security teams.

A useful pilot, then a measured rollout

Scoping identifies one representative workflow and its baseline. The pilot uses approved test material and explicit expected results. Acceptance includes incorrect references, unavailable systems, unauthorised requests and actions that require approval, not only a successful demonstration.

We then agree the production responsibilities: monitoring, incident handling, change review and the person able to suspend the connection. Measure the total time to an accepted result, including human checking and corrections. A faster generated answer is not by itself a productivity gain.

Teams that know how to use their AI

Training is tied to the tools and tasks employees actually use. We cover context preparation, source verification, data boundaries and when to ask a person to decide. Technical teams also need shared repository instructions, test practices and a review process for AI-generated changes.

For managers, the priority is deciding which workflows to authorise and how to evaluate the outcome. For operators, it is knowing what the assistant can do, what it cannot do and how to recover when a request fails. Training and operating documentation are part of adoption, not an optional slide deck after deployment.

Understand before choosing

Our English resources cover platform choices, integration in Morocco and practical software-engineering workflows with Claude Code and OpenAI Codex. They distinguish a proposed pilot from a verified production result, and product documentation from project-specific guarantees.

You can start with a comparison or a maturity diagnostic before requesting a consultation. For a first conversation, bring a task, the software involved, its version and the expected result. Do not send passwords, API keys, payroll exports or confidential customer files through the public contact form.

Define your next AI deployment.

Align the business priority, systems, access and acceptance criteria before commissioning a pilot.

Request a scoping discussion