Berger-Levrault

About the Company

Berger-Levrault is an international software publisher headquartered in France, providing solutions and services to local authorities and administration, healthcare facilities, industries and educational institutions. The Group has offices in Europe, Morocco and Canada, and serves 40,000 customers across 5 continents.

Berger-Levrault is the leading specialist in Human Resources and Financial Management software solutions for public and semi-public sectors in France and Spain, and is also the European leader in maintenance management solutions.

As a long-term, reliable local stakeholder, Berger-Levrault works with its customers and partners to make the world simpler and safer, through a secure digital environment accessible to all.

The Group's 2,200 employees are committed to responsible digital transformation, and work every day to meet their client's social, economic, and environmental challenges by creating trusted digital solutions.

Listed Jobs

Company Name

Berger-Levrault

Job Title

STAGE 2026 - Recherche Opérationnelle et IA - F/H

Job Description

**Job title** 2026 Internship – Operational Research & AI (F/M) **Role Summary** Six‑month, spring‑2026 internship focused on tactical spare‑parts inventory optimization for an industrial maintenance software platform. Apply deterministic inventory models, advanced classification, and modern optimisation techniques (simulation, ML, or multi‑agent) to improve stock levels over 3‑18 month horizons. Deliver a prototype decision‑support demonstrator. **Expectations** - Complete a full internship term (6 months). - Compile deliverables for senior research managers. - Communicate technical findings to multidisciplinary teams. **Key Responsibilities** - Analyse and map CARL Source data linking parts, equipment, and interventions. - Assess demand, lead‑time, and criticality variability. - Perform extended ABC classification incorporating cost, usage, and criticality. - Build a baseline deterministic model (reorder points, safety stock, EOQ). - Conduct sensitivity analysis on stock‑policy parameters. - Design and implement an advanced module (event‑driven simulation, simple ML forecasting, or multi‑agent system) as data diagnostic allows. - Benchmark policies on cost, service level, and resilience. - Produce stock‑level recommendations by category. - Develop and demo a visual decision‑support tool. **Required Skills** - Proficiency in Java and/or Python. - Strong data‑analysis capabilities. - Knowledge of deterministic inventory models (ROP, safety stock, EOQ). - Experience in operational‑research modelling (LP, constraint programming). - Understanding of supply‑chain concepts and Industry 5.0 principles. - Familiarity with maintenance‑management environments or GMAO solutions. - Bonus: simulation, machine‑learning, or multi‑agent system experience. **Required Education & Certifications** - Master’s candidate or final year student in Computer Science, Data Science, Operations Research, or Industrial Engineering. - No specific certifications required.

Labège, France

Hybrid

27-11-2025

Company Name

Berger-Levrault

Job Title

STAGE 2026 - Génie logiciel - Infrastructures Cloud & Simulation énergétique - F/H

Job Description

**Job title** Software Engineering Intern – Cloud Infrastructure & Energy Simulation (Stage 2026) **Role Summary** A six‑month internship focused on designing, developing, and validating a web‑application energy‑consumption simulator that optimizes cloud or on‑premises infrastructure sizing. The intern will apply Green‑IT data, build predictive models, prototype the simulator, and create a scenario‑comparison interface to support sustainable infrastructure decisions. **Expectations** - Enrollment in a Master II (or equivalent) program with a spring 2026 start date. - Strong interest in Green‑IT, digital sobriety, and infrastructure optimization. - Ability to work autonomously, conduct scientific analysis, and collaborate in a multidisciplinary team. **Key Responsibilities** 1. Analyze historical Green‑IT experimental data to extract key metrics for simulation. 2. Define and implement a simulation model that estimates energy consumption, power usage, and carbon footprint from application and infrastructure parameters. 3. Prototype the simulator in a web‑stack environment, producing predictions for CPU load, memory usage, energy usage (kWh), and cloud‑target recommendations. 4. Validate the model against real or prototype applications and refine accuracy. 5. Develop a visualization interface to compare different infrastructure scenarios and generate dimensioning recommendations. 6. Document methodology, results, and best‑practice guidelines for future use. **Required Skills** - Web development in Java, Python, Node.js or equivalent. - Data analysis and modeling: Python, Pandas, SQL, Jupyter notebooks. - Cloud fundamentals (public cloud platforms, virtualization). - Container orchestration: Docker, Kubernetes (basic familiarity preferred). - Monitoring and observability tools: Prometheus (basic familiarity). - Scientific rigor, curiosity, problem‑solving, and clear technical communication. - Team collaboration and self‑directed work. **Required Education & Certifications** - Current Master II (or equivalent) in Software Engineering, Computer Science, or related field. - No mandatory certifications, though knowledge of Green‑IT principles or cloud certifications (e.g., AWS, Azure, GCP fundamentals) is a plus.

Limonest, France

Hybrid

27-11-2025

Company Name

Berger-Levrault

Job Title

STAGE 2026 - IA / NPL - Détection et gestion d'incertitude dans les réponses d'agents IA- F/H

Job Description

**Job Title** AI/ML Internship – Uncertainty Detection in LLM‑Based Conversational Agents (June 2026 – Dec 2026) **Role Summary** Execute a research‑oriented internship aimed at developing and evaluating methods for detecting and quantifying uncertainty in responses from large language model (LLM) agents. The role involves designing reliability‑evaluation modules, prototyping semantic scoring and visual uncertainty feedback, and conducting experimental validation to enhance the safety and trustworthiness of an enterprise AI assistant. **Expectations** - Deliver an analytical report comparing existing uncertainty estimation and confidence‑calibration techniques for LLMs. - Produce a prototype agent that scores answer reliability, visualizes confidence, and optionally offers corrective suggestions. - Develop an experimental protocol and publish a scientific report on precision, robustness, and user perception. - (Optional) Advise on integration into the Athena AI suite within the organization. **Key Responsibilities** - Review literature on probabilistic estimation, self‑consistency, chain‑of‑thought scoring, contradiction detection, and semantic similarity metrics (SBERT, cosine distance). - Design a reliability‑assessment module using semantic similarity, inter‑agent coherence, and external source verification (web APIs, corporate corpora). - Implement a semantic scoring system to identify internal contradictions and logical divergences. - Build a user‑interface prototype (e.g., FastAPI/Streamlit) that displays confidence (bars, colors, reformulations) and collects user feedback. - Set up and run quantitative experiments measuring accuracy, calibration, robustness, and user trust. - Document methods, results, and recommendations for production deployment. **Required Skills** - Proficiency in Python; experience with LLM integration (Mistral‑AI, Hugging Face). - Familiarity with NLP frameworks (LangChain, SBERT, Weaviate or similar vector databases). - Knowledge of empirical evaluation: statistical analysis, hypothesis testing, experimental design. - Ability to prototype web services (FastAPI, Streamlit) and visualize metrics. - Strong written and oral communication for scientific reporting. - Curiosity for experimentation and methodological rigor. **Required Education & Certifications** - Current enrolment in a Master’s or equivalent graduate program in Artificial Intelligence, Data Science, NLP, or Computer Science. - Demonstrated coursework or projects involving language models, embeddings, or uncertainty estimation. ---

Labège, France

Hybrid

08-12-2025

Company Name

Berger-Levrault

Job Title

STAGE 2026 - Génie logiciel - Comprendre un framework par rétro-ingénierie semi-automatique - F/H

Job Description

Job title: Software Engineering Internship – 2026: Framework Reverse Engineering Role Summary: A 6‑month internship focused on developing semi‑automatic reverse engineering methods to map and analyze data flows within legacy software frameworks. The role involves literature review, prototype implementation, and data visualization to aid maintainability and comprehension. Expectations: Analyze complex, poorly documented frameworks; produce scientific state‑of‑the‑art review; design and implement prototypes for data‑flow detection; synthesize findings and identify research challenges; work autonomously within a research‑engineering team. Key Responsibilities: - Conduct comprehensive literature review on data‑flow mapping and software cartography. - Develop prototypes that extract and visualize dependency graphs, architectural patterns, and dynamic execution traces. - Apply static and dynamic analysis techniques to real software artifacts. - Document methods, results, and scientific insights. - Present findings to research and engineering stakeholders. Required Skills: - Proficiency in an object‑oriented language (Java, C#, Pharo, etc.). - Experience with Git version control. - Ability to write clear, concise technical reports. - Strong analytical and research skills (literature review, scientific method). - Familiarity with software visualization tools (e.g., Understand, CodeCity, industry literature on graph extraction). - Autonomy and self‑management of research tasks. Required Education & Certifications: - Current enrolment in the final year of a Master’s or engineering degree in Software Engineering, Computer Science, or equivalent. - Coursework or experience in static/dynamic analysis, software architecture, or reverse engineering is preferred. ---

Labège, France

Hybrid

09-12-2025