Freelancer profile translated to English.
Back to original languageAbout El Hadji Abdoulaye
Senior AI Engineer & Forward Deployed Engineer (FDE) — Expert
AI Generative with 7+ years of experience. I design and deploy
RAG architectures, autonomous AI agents, and GenAI backends
directly within the infrastructure of large accounts (AXA,
Crédit Agricole, Publicis).
My FDE specialty: transforming complex business needs
into production AI applications — from POC to MVP in weeks,
not months.
🎯 Proven results
- -60% support time (RAG chatbot, Crédit Agricole)
- -35% detected fraud (ML, Reliance HMO)
- ML/LLM learning path in 50+ AXA entities
- 1000+ users/day on my RAG systems
🤖 My 5 FDE expertise areas
1. Rapid Prototyping POC → MVP → Production
Direct deployment in client infrastructure, backend logic, API
integration, automated workflows, rapid iterations.
2. RAG & Document Search
Scalable RAG, multimodal (text, images, graphs), semantic
search on 10,000+ docs, reranking, evaluation.
3. AI Agents & Multi-Agent Frameworks
Autonomous agents (Pydantic AI, LangGraph, MCP), multi-agent
orchestration, LLM orchestration, agentic AI.
4. GenAI Backend & LLM APIs
LLM APIs (FastAPI), API integration, token management,
guardrails, multi-providers (Claude, OpenAI, Bedrock, Azure).
5. MLOps, LLMOps & Fraud Detection
CI/CD, Docker, Kubernetes, LLM monitoring, banking/document
fraud, credit scoring, churn prediction.
🚀 Why choose me
- Battle-tested FDE: POCs deployed in real production
- AI Engineer + trainer (250+ OpenClassrooms reviews,
AXA workshops in 50+ countries)
- Large accounts: AXA, Crédit Agricole, Publicis
- International: Senegal, Nigeria, Switzerland, France
🛠 Stack
GenAI: Claude, OpenAI, LangChain, LangGraph, MCP, RAG, LangSmith, Guardrails
Backend: FastAPI, Python, API Integration
MLOps: Docker, Kubernetes, GitHub Actions, Terraform
Cloud: AWS (Bedrock), Azure, GCP
📞 Need an FDE to transform a business need into a production
AI app? Let's chat for 15 min.
French
Native or bilingual
English
Fluent
Can work on-site
Paris (up to 50km)
Experience
- AXA Group Operations - GDAI FRANCE
On Malt
AI Engineer — ML/LLM Engineering & GenAI IndustrializationBANKING AND INSURANCENovember 2025 - March 2026 (3 months)Paris, FranceAssignment within the AI Engineering project of AXA's GDAI(Group Data & AI) division, responsible for defining and deployingML/LLM Engineering best practices across the global AXA group (entities in 50+ countries).- Industrialization of ML/GenAI pipelines on AzureML: automation of model training, packaging, and deployment to production.
- Design and industrialization of a multi-agent system for automated insurance claims processing — dossier triage, document extraction, fraud signals, decision support: deployed on Azure AI Foundry with LangGraph orchestration.
- Migration of the AI agents platform to Azure AI Foundry: porting existing workflows, native integration with Azure AI Search (hybrid search) and Azure OpenAI Service.
- Implementation of MCP servers (FastMCP) for standardized exposure of business tools to LLM agents: standardization of access to insurance APIs across entities.
- Establishment of robust MLOps and LLMOps workflows: GitHub Actions / Azure DevOps CI/CD, MLflow experiment management, monitoring and observability (Langfuse), automated evaluation (DeepEval).
- Design and delivery of the group-wide ML / MLOps / LLMOps training program: modules on best practices (prompt engineering, evaluation, monitoring, AI agents), practical workshops for international entities.
- Multi-cloud evolution of the training platform: migration of ML Engineering tracks to Azure (AzureML) and Databricks; harmonization of best practices.
- Technical support to entities for industrializing AI POC/MVP: architecture, CI/CD, scalability, security, and compliance.
Stack:Azure (AzureML, Azure AI Foundry, Azure AI Search, Azure OpenAI, Azure Functions, Azure DevOps), Databricks, Python, MLflow, GitHub Actions, Docker, Kubernetes, LangChain, LangGraph, MCP (FastMCP), A2A, Langfuse, DeepEval - CREDIT AGRICOLE SASenior AI Engineer ConsultantBANKING AND INSURANCESeptember 2024 - Today (1 year and 11 months)Montrouge, FranceProject: Development of a RAG platform for industrializing document search use cases.Tasks:
- Development of 2 core Python libraries for the LLM wrapper platform: Multi-provider abstraction and semantic search: Configurable indexing and search pipelines (chunking, search, reranking)
- Integration of a feature for managing abbreviations and technical glossary specific to the business.
- Implementation of a rate limiting system: mechanism to prioritize certain uses and bypass Bedrock restrictions regarding the number of requests and tokens per minute.
- Marketing content generation: Implementation of an LLM module capable of automatically creating content for products featured on the website.
- Implementing automatic and manual approaches for evaluating RAG pipelines.
- Integration of multimodality to process and understand images and graphs within the RAG system.
- Participating in scoping workshops to define the needs for new use cases utilizing the RAG tool's assets.
- Orchestration of AI agents with AWS Lambda step functions.
- Participating in the deployment of new use cases.
- Technologies: AWS (Bedrock, Sagemaker, SQS, S3, Lambda...) - Python - FastAPI - Memcached, Postgres pgvector, LLM, Prompting, GitLab CI, Docker, Kubernetes
- Publicis Re:sourcesSenior Data Science Consultant GEN AIHUMAN RESOURCESFebruary 2024 - August 2024 (6 months)Paris, France
Financial AI Matching System
● Architected deep learning solution for automated invoice-to-purchase order matching withadvanced scoring mechanisms● Improved model performance through feature engineering and neural network optimization● Implemented MLOps pipeline with continuous deployment using Azure DevOpsHR GenAI Chatbot
● Built end-to-end RAG chatbot for HR inquiries with multi-format document processing (PDF,Word, Excel)● Developed generic data loader using nlm-ingestor for various document types● Implemented microservices architecture with FastAPI and deployed on Azure
Recommendations
Be the first to recommend El Hadji Abdoulaye
Help this freelancer shine by sharing your experience working together.
These freelancer profiles also match your criteria
Agatha Frydrych
Backend Java Software Engineer
4.7
(3)
2
Baptiste Duhen
Fullstack developer
4.6
(4)
5
Amed Hamou
Senior Lead Developer
4
(2)
7
Audrey Champion
Web developer
4.3
(3)
4
Education
- Engineering Degree in Computer Science and TelecommunicationsEcole Polytechnique de Thies2018
Certifications
- Professional Machine Learning EngineerGoogle Cloud