Amel Ould Amara

Designed and delivered production-ready AI solutions that optimize industrial operations using GenAI, Agentic AI, speech technologies, and MLOps on Microsoft Azure.

- Led the design and deployment of an Agentic AI voice assistant replacing manual intervention reporting with intuitive voice interactions.

- Architected a conversational system combining Speech-to-Text, LLMs, and enterprise knowledge to support industrial operations in real time.

- Delivered a production-grade Azure architecture with automated deployment, monitoring, and CI/CD pipelines.

- Led the design and deployment of GenAI-based Agentic workflow on Microsoft Azure to automate freight train operations management, replacing manual, paper based tracking with real-time digital systems.

- Built intelligent agents capable of extracting train composition, wagons, cargo information, and customer destinations, automatically integrating results into enterprise databases and reporting systems.

-Led the design and deployment of a GenAI assistant enabling operators to query freight train data using natural language.

- Built a Text-to-SQL architecture over Oracle databases to translate user requests into SQL queries minimizing LLM exposure to sensitive data while reducing inference costs.

- Ensured data confidentiality by retrieving only the relevant information required to answer user queries, avoiding exposure of sensitive data to LLMs.

- Built anomaly detection models in Python to identify irregular patterns in geological datasets.

- Designed data pipelines on Fabric Lakehouse for data processing.

key words: Agentic AI, GenAI, LLMs, RAG, LangChain, MLOps ,Microsoft Azure, Azure AI Services, OpenAI API, Python, SQL, Microsoft Fabric, Azure DevOps, CI/CD, Pydantic, Oracle

Designed and delivered AI , Machine Learning and GenAI solutions supportinghealthcareandpharmaceuticaldata products, from model development to production deployment.

- Designed and Developed advanced forecasting and recommendation models to improve decision-making and optimize medical interactions.

- Built multimodal analytics solutions combining structured data, text, and document intelligence to generate business KPIs using NLP and Large Language Models (OpenAI API).

- Designed a GenAI-powered document processing solution using LLMs to extract and validate information from medical prescriptions and healthcare forms, identifying missing, incomplete, or inconsistent data to improve data quality and operational efficiency.

- Managed cloud-based data environments on AWS, leveraging Amazon S3, Redshift, and SageMaker for scalable data processing and model deployment.

- Data Pipeline Orchestration : Configuring data processing pipelines and batch forecasting workflows on Databricks (Spark, Airflow, MLflow).

- MLOps : Providing technical support and maintaining deployed models (CI/CD, AWS, Airflow, Kedro, Kubernetes)

Technologies:Python , Generative AI , LLMs , OpenAI API , NLP , AWS (S3, Redshift, SageMaker) , Databricks , Spark , Airflow , MLflow , Kedro , Kubernetes , CI/CD

- Founded and led a consulting team to deliver an AI-driven innovation project for Safran, addressing sustainability challenges in the aviation sector.

- Recruited and coordinated a cross-functional team, managing project planning, task allocation, and client communications.

- Designed an end-to-end AI solution leveraging Computer Vision and Machine Learning to reduce onboard waste and improve operational sustainability.

- Directed data collection, analysis, and model development activities, transforming operational data into actionable business recommendations.

- Delivered strategic and technical recommendations to Safran, demonstrating the feasibility and business impact of AI-driven waste reduction initiatives.