Clicky
Skip to content

Curriculum Vitae

Antoine Guillaume

Email: antoine.guillaume45@gmail.com
GitHub: github.com/baraline
LinkedIn: antoine-guillaume
Born: December, 1994

Profile

AI engineer and researcher, PhD in Computer Science and Machine Learning, with 8 years of software engineering experience — 5 of them designing, deploying and evaluating AI systems in production.

My current work focuses on generative AI over unstructured data (RAG, semantic search, information extraction from document corpora) and on rigorous model evaluation: choosing business-relevant metrics and weighing cost/performance trade-offs between open-source and proprietary models. Before that, I worked on industrial machine learning — predictive maintenance, rare event prediction, forecasting — and on the industrialization of AI projects. I also lead training and change management workshops for these types of projects and provide consulting services to assist in the creation of enterprise AI platforms, and I participate in the open-source AI ecosystem through NumFOCUS with aeon.

Employment History

AI Engineer

Novahé & Constellation | 2024 – Present

Technical lead on the development, production deployment and evaluation of generative AI solutions, in collaboration with IBM partners.

  • Designed and deployed LLM-powered features over document corpora: a RAG pipeline connected to SharePoint, a Wiki and the ITSM history to retrieve procedures and client-specific information, with automatic procedure creation for cases not yet documented in the knowledge base
  • Built the semantic search layer: LLM and embedding models served through IBM watsonx.ai, populating Milvus vector databases
  • Evaluated and selected models, tools and frameworks (open-source and proprietary) based on cost/performance trade-offs and measured quality on real use cases
  • Developed an AI agent platform orchestrating pipelines, connected to Azure for access and rights management and to IBM watsonx Orchestrate to import specialized agents
  • Built and maintained a library of skills and MCP servers, made available to users through a Claude Enterprise deployment
  • Supervision and structuring of the group's R&D activities, including Research Tax Credit applications
  • Participation in internal and external AI projects in collaboration with partners (IBM, Oracle)
  • Training activities on AI and risk prevention
  • Active participation in the industrialization of projects, using technologies such as Docker, Kafka, Prometheus and Grafana, deployed on Red Hat systems and interfaced with the Azure environment
  • Pre-sales and consulting activities: use-case scoping and presentation of AI solutions to clients and partners

Key Skills: Python, Generative AI, RAG, Vector Databases, AI Agents, Model Evaluation, Data Engineering, Project Management, Communication

Postdoctoral Researcher

ENSTA Paris | 2022 – 2024

  • Development of rare event and anomaly detection models from sensor and operational data for an EDF use case (Python + Numba, ETL, Docker)
  • Extensive work on evaluation frameworks: selecting metrics that reflect the real business value of predictions rather than statistical performance alone
  • Collection and preprocessing of datasets from technical documentation, with feature engineering steps turning heterogeneous, high-volume sources into usable datasets
  • Proposal of algorithmic solutions adapted to business constraints, aligned with EDF domain experts
  • Communication around the results and the method in the group's seminars and stakeholder meetings

Key Skills: Python, Numba, Data Engineering, Predictive Modelling, Model Evaluation, HPC, Communication

Part-time Lecturer

University of Orléans | 2019 – 2022

  • Teaching activities for bachelor's and master's degree students in statistics and machine learning applied to time series
  • Educational content creation

Research and Development Engineer

Worldline, Financial Services Department | 2018 – 2022

  • Project management and development of predictive maintenance and anomaly detection solutions for fleets of automated teller machines, integrating research results into deployed applications
  • Development of new predictive models and data pipelines
  • Development of a REST API and backend components to integrate the models into an existing application and into reporting tools
  • Built and maintained data pipelines and dashboards in Java, SQL and Splunk for operational monitoring and data quality

Key Skills: Java, SQL, Splunk, Predictive Modelling, API, Project Management, Dashboards

Education

Ph.D. in Computer Science & Machine Learning

University of Orléans | 2019 – 2022

Thesis: Time series classification with Shapelets: Application to predictive maintenance

Research Fields: Time series analysis, transformation and classification, anomaly detection, survival analysis, trigger systems

Additional Activities: - Member of the organization committee of the yearly Ph.D. conference of the doctoral school - Ph.D. Students representative at the council of the doctoral school

Key Technologies: Python, C++, HPC, Scientific Writing, Communication

M.Sc. in Computer Science

University of Orléans | 2016 – 2018

Specialization: Intelligent and Secure Mobile Computing (IMIS)

Additional Diploma: Hadoop and Chemoinformatics via the Orléans Numérique "GSON" Graduate School

End-of-study Internship: (6 months) Development of a predictive maintenance tool for ATMs at Worldline in Blois

B.Sc. in Computer Science

University of Orléans | 2012 – 2016

Specialization: Software Engineering

End-of-study Internship: (4 months) Development of a network simulation and analysis module in Perl for an existing application at Sopra Steria in Orléans

Technical Skills

Programming Languages: Python, Java, C++, SQL, Bash, Perl

Generative AI & LLMs: RAG architectures, vector databases (Milvus), semantic search, LLMs and embeddings, AI agents (Langchain, Langgraph, Langflow/n8n), MCP servers, information extraction from heterogeneous documents

Model Evaluation & Reliability: evaluation frameworks, selection of business-relevant metrics, cost/performance trade-offs between open-source and proprietary models, testing and validation, reproducibility, data quality

Machine Learning & Data Science: NLP, deep learning, feature engineering, time series analysis, predictive modelling, anomaly detection, rare event prediction, survival analysis

Data Engineering: Kafka, ETL and data pipelines, PostgreSQL / SQL Server, training and inference pipelines, HPC, Numba

DevOps & Infrastructure: Docker, Kubernetes, Prometheus, Grafana, Red Hat, Azure, IBM Cloud / watsonx, CI/CD

Tools & Frameworks: Git, API development, dashboards, code quality, TDD and automated testing

Leadership & Communication: technical mentoring, AI roadmap and use-case scoping, workshops and training, communication to non-technical audiences, cross-functional collaboration, consulting and pre-sales

Open Source Contributions

aeon-toolkit - Core Developer

NumFOCUS | 2022 – Present

  • Implementation of algorithms from the time series literature with associated tests and documentation
  • Supervision of contributors and interns during Google Summer of Code
  • Repository: github.com/aeon-toolkit/aeon

Key Technologies: Python, Code Quality, Documentation, Testing, CI/CD, Networking

Research Interests

  • Time Series Analysis and Classification
  • Generative AI over Unstructured Data (RAG, Semantic Search, Information Extraction)
  • Evaluation of AI Systems
  • Predictive Maintenance
  • Rare Event Prediction
  • Machine Learning for Industrial Applications
  • Anomaly Detection

Certifications

  • IBM Cloud — sales and technical level (watsonx.ai, watsonx.data, watsonx Orchestrate)

Languages

  • French: Native
  • English: Fluent

Last updated: August 2026