Hi! I'm Mehdi
Hayani Mechkouri

AI Engineer & Problem Solver

I engineer AI agents that democratize data — turning complex databases and documents into conversational interfaces that anyone in an enterprise can use. Currently building at Saint-Gobain, winning awards doing it.

8+Years Python Exp.
2025GenAI @ Scale Award
20+Business Units Served
FAME Competition - Valeo AI Solution

"Solving the automotive data bottleneck with Generative AI for Valeo at the FAME Innovation Forum."

Current MissionData Scientist x AI Engineer @ Saint-Gobain · Since Sept 2024

Empowering the Lead & Grow Strategy

We build specialized AI agents to serve the company's 2030 roadmap — transforming how we leverage data, automate workflows, and enable a continuous customer obsession mindset across 80 countries.

Better Data Democratisation

Eliminating bottlenecks with NL-to-SQL architectures. Giving global business units direct, self-serve access to their production data without engineering dependency.

Tasks Automatisation

Designing and deploying secure multi-agent systems to streamline enterprise workflows. Full technical ownership from system architecture to production CI/CD.

Customer Obsession Mindset

Enabling continuous improvement by building intelligent pipelines that turn complex operational streams and client feedback into actionable voice-of-customer insights.

Professional Projects

Scalable AI systems designed to solve real-world enterprise bottlenecks.

Award Winning Live Production

Query Genie

Bridging the gap between human intent and complex databases. I engineered specialised semantic agents using the Model Context Protocol (MCP) to search datasets, navigate data models, and execute natural language queries directly on Snowflake — no SQL required.

The Problem

Data was locked behind engineering queues. Every business question required a custom SQL request, consuming 80% of data team bandwidth.

The Result

98% NL-to-SQL accuracy on production Snowflake schemas. Eliminated the bottleneck for 20+ global business units. Won the Saint-Gobain GenAI @ Scale Award 2025.

The Scale

Deployed across Saint-Gobain globally, serving business stakeholders who now self-serve their data without any engineering dependency.

MCPSnowflakeAzure OpenAIFastAPICollibraPostgreSQLDatadogAzure DevOps

AI4Legal

Multi-Agent

Automated legal contract analysis and AI-powered redlining for enterprise compliance teams. A multi-agent pipeline reads, flags, and annotates NDAs and service agreements — reducing manual review from 5 hours to under 30 minutes with zero critical clause misses. Deployed on Azure with full Datadog observability.

Multi-AgentAzure OpenAIFastAPIDatadog

VENTURI

Voice intelligence pipeline for call-centre analytics. Speaker diarisation separates agent from customer speech, enabling automated QA scoring, feedback categorisation, and trend detection at scale — turning raw audio into actionable insight.

DiarizationPythonSpeech AI

Personal Projects

Hackathons, side experiments, and personal engineering challenges beyond the enterprise walls.

Datacard

"GenAI-enhanced cross-platform mobile app."

Overview

Graduation capstone project at Arts et Métiers ParisTech. A cross-platform React Native app that reads smartphone sensor data — accelerometer, gyroscope, and ambient metrics — and feeds them as creative constraints to a GPT-powered brainstorming engine, turning physical motion into ideation prompts.

React NativeJavaScriptOpenAI APIPrompt EngineeringMobile Development
Learning Path

Education & Certifications

Formal training and professional accreditations

Universities & Engineering Degrees

Arts et Métiers ParisTech — École Nationale Supérieure d'Arts et Métiers
Degree

Engineering Degree — Digital Transformation of Enterprises

Specialised in Digital Transformation encompassing Data & AI, IoT, Cloud Computing, and Agile. Thesis: built 'Datacard', a GenAI-enhanced mobile app merging smartphone sensor data with GPT-powered ideation. Applied genetic algorithms to optimise high-pressure hydrogen reservoir geometries.

Arts et Métiers ParisTech — École Nationale Supérieure d'Arts et Métiers

Sept 2022 - Aug 2024

ENSAM Meknès — École Nationale Supérieure d'Arts et Métiers
Degree

Engineering Degree — Industrial Engineering, AI & Data Science

Six-year engineering degree spanning Industrial Engineering, Data Science, and AI. Co-inventor of Patent MA57031A1 — an intelligent real-time behavioural analysis system for driver and workplace safety using computer vision. Applied Lean Six Sigma, Industry 4.0, and Digital Twin methodologies throughout.

ENSAM Meknès — École Nationale Supérieure d'Arts et Métiers

Sept 2018 - Aug 2024

Achievements

Competitions & Awards

Recognitions for algorithmic problem-solving, robotics, and AI innovation.

5Competitions
4Podium Finishes
3Domains
WinnerSaint-Gobain Digital & IT
Jan 2025

Saint-Gobain GenAI @ Scale Award

Awarded by Saint-Gobain Digital & IT in 2025 for driving the most impactful Generative AI initiative across all global business units. Recognised for architecting QueryGenie, an enterprise NL-to-SQL system serving 20+ global BUs.

Media Gallery
Saint-Gobain GenAI @ Scale Award photo
Redirect
3rd PlaceForum Arts et Métiers
Nov 2023

FAME Innovation Competition

Secured 3rd place for delivering scalable Generative AI solutions for Valeo — tackling real operational challenges in the automotive industry. Our project addressed production data scarcity, driver safety, and documentation access, all while prioritising energy-conscious AI design.

  • Generated synthetic defect data using GANs to overcome limited production data.
  • Built an AI-driven sound generator to enhance driver comfort and pedestrian awareness.
  • Developed a RAG assistant to streamline access to technical manufacturing documentation.
Media Gallery
FAME Innovation Competition photo
Expand
FAME Innovation Competition photo
Expand
3rd PlaceArts & Métiers Meknès
May 2021

African Robotic Brains

Awarded 3rd place for engineering an autonomous robot capable of solving complex mazes. I led the conception and implementation, integrating advanced sensor processing, pathfinding algorithms, and hardware control.

  • Implemented a graph-based mapping algorithm for autonomous maze exploration.
  • Processed convolutional path data using infrared sensors.
  • Integrated PID control for precise motor actuation and an RFID module using C++ on an Arduino Mega.
Media Gallery
African Robotic Brains photo
Expand
2nd PlaceUniversité Privée de Fès
May 2021

UPF Coding Challenge

Secured 2nd place in a high-performance competitive programming challenge evaluating mathematical, graph, and dynamic programming skills. Competing in a dual team format, we demonstrated precision algorithmic problem-solving—finishing over 30 minutes ahead of the 3rd place team.

Media Gallery
UPF Coding Challenge photo
Expand
CompetitorACM ICPC
Dec 2020

ACM MCPC 2020

Competed in the Moroccan Collegiate Programming Contest (MCPC) — Morocco's most prestigious ACM-ICPC national championship, gathering the top university teams from across the country. Utilising C++, our team solved 4 out of 10 demanding algorithmic problems under strict time and memory constraints, showcasing rigorous team coordination and technical mastery.

Media Gallery
No images
available

The Arsenal

Architectural frameworks for agentic intelligence and scalable enterprise systems.

My Superpower

Vibe Coding & AI Velocity

I master the foundational concepts of software engineering and system architecture, allowing me to leverage AI tools like Claude Code and Antigravity as velocity multipliers. While AI handles rapid implementation, I remain focused on high-level design, problem-solving, and quality assurance.

Claude CodeAntigravityCursorVibe CodingAgentic Development

Intelligence & Orchestration

AIAutonomous AgentsDiarizationFoundation ModelsGenerative AIInsight SynthesisLLM OrchestrationLLM-as-judgeLLMsMCPMachine LearningMulti-Agent SystemsOpenAI APIPrompt EngineeringRAGVertex AI

Languages

PythonTypeScriptJavaScriptSQLC++Bash

Data & Cloud Architecture

AWSAzureAzure DevOpsAzure OpenAICloud ComputingCollibraData AnalysisData PrivacyData VisualizationDatadogETL PipelinesGCP Cloud RunGeoJSONMongoDBPostgreSQLSQLSnowflakeSnowflake APIVector Databases

Application Engineering

A/B TestingDocx XML formatFastAPIGoogle ADKJavaScriptMobile DevelopmentNext.jsPythonReact NativeStreamlitTypeScriptuvPydantic

Systems & Reliability

AutomationBicepC++CI/CDDockerGitHub ActionsJfrog XrayPerformance OptimizationSystem Architecture

Ready to build intelligent systems that scale? Let's connect.

Fastest response via LinkedIn, or use the form below.

Success

Message sent successfully! I will get back to you soon.