Agentic AI · Product Strategy · End-to-End AI Development
From conversations to commercial decisions.
Designed and built an AI-powered commercial intelligence system that transforms unstructured property-buyer conversations into qualified opportunities, prioritised leads, and recommended sales actions.
- Role
- Founder · AI Product Architect · End-to-End Builder
- Sector
- Real-estate technology
- Scope
- Agentic AI · production MVP
Founder-built AI product · Production MVP · Live demonstration
01 · The challenge
Real-estate developers receive inbound enquiries from buyers with different budgets, financing situations, purchase timelines, and levels of intent. Sales teams often spend significant time manually interpreting conversations, identifying serious buyers, and deciding whom to follow up with.
The problem wasn't a lack of leads. It was a lack of actionable intelligence.
02 · The solution
Built RIADIS.ai, an agentic AI system that combines conversational AI, buyer-signal extraction, qualification logic, automated follow-up and recommended next actions to turn buyer conversations into structured commercial decisions.
From incoming conversation to prioritised sales opportunity.
03 · How it works
01 — Engage
An AI conversational agent handles inbound buyer conversations in French and Moroccan Darija, gathering relevant information about the buyer's requirements.
02 — Understand
The system extracts and structures commercial signals, including budget, financing readiness, property preferences, purchase timeline, and engagement.
03 — Qualify
A qualification engine evaluates buyer signals and assigns a priority score from 0–100, with explainable priority tiers.
04 — Act
The system recommends the next commercial action, supports follow-up workflows, and surfaces high-priority opportunities in a sales cockpit.
04 · In practice
From one conversation to a qualified opportunity.
A buyer chat on WhatsApp becomes a scored lead with budget, timeline and a recommended next action.
Yasmine El Amrani
online
WhatsApp Business
RIADIS replies in French and Darija · escalated to the team on strong signals
RIADIS
Hello Yasmine. We still have 3 two-bedroom units within that budget. What timeline are you buying on?
2:28 PM
I want to buy before the end of the month. I have a 40% down payment and I'll be in Marrakech on Monday.
2:31 PM
RIADIS
Great. I can arrange a viewing on Monday. Would 2 PM work?
2:31 PM
Yasmine El Amrani
🇫🇷 Paris · MRE
86/100
- Budget
- 1.8M MAD
- Type
- 2-bed apartment
- Timeline
- < 30 days
05 · System architecture
Input
Inbound buyer conversations
WhatsApp / digital enquiries
Understanding layer
Conversational AI
Engagement and context
Signal extraction
Structured buyer intelligence
Decision core
Qualification & decision engine
Scoring · Priority tiers · Next-best action
Commercial action
Sales cockpit
Prioritised opportunities
Follow-up workflows
Alerts and engagement
06 · Live demo
Don't just read about the AI. Talk to it.
See the architecture in action.
Experience RIADIS.ai through a live WhatsApp demonstration configured for Majorelle Avenue, a real-estate project in Marrakech. Interact with the AI agent as a prospective buyer and explore how it gathers information used to qualify commercial opportunities.
Try the Live AI AgentLive AI demonstration · Majorelle Avenue · Opens WhatsApp
RIADIS · Majorelle Avenue
- Hello, I'm interested in Majorelle Avenue in Marrakech.
- Welcome! Are you looking for a home to live in or an investment?
- To live in — a 2-bedroom, ideally this year.
- Noted. What budget range are you considering?
07 · My role
Founder · AI Product Architect · End-to-End Builder
- Identified and validated the commercial problem through developer and buyer discovery.
- Defined the AI product strategy, system architecture, and commercial qualification framework.
- Built and integrated the AI conversational workflows, frontend, backend, and database.
- Developed the buyer-signal extraction, lead scoring, follow-up, and prioritisation logic.
- Designed the commercial cockpit to translate AI analysis into actionable sales decisions.
- Took the product from concept to a functioning production MVP through AI-assisted development.
08 · Evidence of execution
- Buyer prospects processed
- 399
- Priority opportunities
- 77
- Property projects
- 15
The platform demonstrates how unstructured buyer conversations can be converted into structured commercial intelligence and actionable sales priorities.
Platform figures are operational metrics, not independently verified sales conversions or revenue impact.
09 · What this demonstrates
The ability to take an applied AI opportunity from customer discovery through product architecture, implementation, and operational deployment — connecting AI capabilities directly to business workflows.