Nacim M.

Producing content regularly, across multiple channels, while maintaining brand consistency: this is the daily challenge for any marketing team aiming for visibility. Iradia Studio addresses exactly this.

I built Iradia's content generation module based on a simple observation: a well-defined topic should be enough to feed all of a brand's channels. No need for an hour of writing per format, no need to manually rephrase for each network.

The Studio generates 5 optimized formats from a single topic with one click: SEO-ready blog post, high-engagement LinkedIn post, structured YouTube script, impactful Reel/Short script, and Instagram caption with a hook. Each format respects the specific codes of its platform: length, tone, structure, call to action.

The whole system is organized in the Planning tool, a drag-and-drop editorial calendar with weekly and monthly views. Generated content is directly placed there. 9 content pillars help structure the strategy by theme, and the platform suggests related ideas based on monitored content.

What changes concretely: the time from 'idea' to 'ready-to-publish content' goes from 3 days to a few hours. An SME without a content team can maintain a regular presence on 5 channels without dedicating their week.

I designed the architecture for growing teams: multi-user, natively multilingual (FR/EN/AR), with fine-grained rights management and usage tracking. Everything integrates seamlessly into the complete Iradia pipeline: monitoring, SEO insights, generation, planning, without friction between modules.

Deployed in production with Stripe, subscription management, admin dashboard, and usage monitoring.

60% of customers leave if their message goes unanswered for more than 5 minutes. For SMEs, restaurants, clinics, real estate agencies, and beauty salons, WhatsApp is the primary sales channel.

However, responding 24/7, in Arabic, English, and French, is beyond the reach of a small team.

Natiq solves this.

I designed and developed Natiq from A to Z: a SaaS platform that connects an AI assistant directly to a WhatsApp Business number. The assistant learns the client's business by ingesting their documents (catalog, FAQ, service guide), then responds to customer messages in real-time – in under 2 seconds – regardless of the language used.

Language detection is automatic. A customer writes in Gulf Arabic, the assistant responds in Arabic. They switch to English, the assistant follows. No manual configuration, no routing.

Onboarding is deliberately ultra-simple: scan a QR code, upload business documents, launch. Less than 5 minutes to have an operational assistant. This choice is deliberate; the target is an SME manager, not an IT director.

They must be able to deploy it themselves, without support.

The dashboard allows real-time monitoring of all conversations, manual takeover of an exchange, and measurement of key indicators: average response time, volume handled, resolution rate.

Agents continuously search and identify profiles across all available platforms and internal databases – without human intervention. What used to take several hours of manual research per day now happens in the background, 24/7. The firm receives qualified profiles without its consultants needing to spend time on prospecting tasks.

Each incoming CV is analyzed, structured, and automatically scored according to the specific criteria for each assignment. No more manual sorting of applications: consultants only intervene on pre-qualified profiles. The processing time for a file goes from several minutes to a few seconds.

Agents track each candidate's status and trigger follow-ups at the right time, with the right message. No candidate falls through the cracks. The response rate increases without the team having to manage a single follow-up manually.

A job description written in the platform is automatically published on LinkedIn, Google, and other channels in seconds. What required manual posting on each platform becomes a single, instant action.

Every interaction, received email, added document, updated status is automatically integrated into the database.

Agents delegate tasks to each other based on their specialization: one agent searches, another analyzes, another follows up. The system scales frictionlessly.