Contract type Fixed-term contract
Duration 18 months
Location Nancy
Start date October - November 2026
Salary 2500 euros (net salary - charges deducted)
Required Education Master's degree (or equivalent) minimum — PhD preferred
Context: This offer is part of the University of Lorraine's interdisciplinary program “B4B” (Biomolecules for the Bioeconomy). Designed to accelerate innovation in the discovery, production, and formulation of biomolecules for key sectors such as agri-food, nutrition and health, and cosmetics, it aims to reconcile the development of biobased molecules with the sustainable management of natural resources and environmental preservation. The recruited research engineer will work specifically on the ECIA-CAB project, which aims to design, using Artificial Intelligence (AI) agents, a fertilizer encapsulation methodology that limits the societal impacts (environmental, social, and economic) of European agriculture.
Within a multidisciplinary framework involving computer scientists, AI and data science specialists from LORIA, management science researchers from CEREFIGE, and agri-food physical chemists from LIBio, we seek to combine different approaches in the automatic extraction of scientific knowledge, the collection of open environmental data, and advanced analysis using machine learning.
The recruited engineer will be at the heart of this project, developing AI agents capable of automating scientific monitoring and enriching predictive models with expert knowledge.
Project Summary
This project aims to reduce the environmental impact of European agriculture through cost-effective encapsulation solutions for inputs, thereby enabling the development of precision agriculture. It will leverage the expertise of the B4B program and artificial intelligence to reduce fertilizer and pesticide use, mitigate health and environmental risks, lower costs, and increase farmers' incomes. The ECIA-CAB project will: (i) develop an AI-based application to determine the optimal composition of the encapsulation matrix based on the characteristics of the biomolecules, as well as local climate and soil data (precision agriculture); (ii) reduce pesticide and phosphate inputs by up to 50% through the use of controlled-release systems (encapsulation); and (iii) decrease costs for farmers by 15% by reducing the amount of active inputs required (minimizing losses to soil and water). (iv) to support the creation of a start-up (software and mobile application); and (v) to reduce the overall impact of agriculture on citizens. By developing input encapsulation, the project aims to provide affordable bioencapsulation solutions enabling reduced, efficient, and sustainable use of inputs for precision agriculture.
Main Tasks
1 — Development of an AI Agent for scientific monitoring, design and deployment of an autonomous AI agent for extracting information from scientific publications:
- Encapsulation formulas: combinations of chemical compounds, ratios, and manufacturing processes.
- Physicochemical properties: stability, solubility, bioavailability, efficacy.
- Environmental impact: ecotoxicity, biodegradability, carbon footprint.
- Economic data: production costs and prices mentioned in the literature.
- Methodologies and criteria for assessing sustainability: circumstantial life cycle analysis, assessment frameworks.
- Technical skills: LangChain/LangGraph, PubMed/ArXiv/Semantic Scholar API, PDF parsing, RAG, Excel/database export.
2 — Open Data Collection and Integration, Development of Open Environmental and Agronomic Data Collection Pipelines (France / Europe)
- Meteorological Data: ERA5, Météo-France, Copernicus
- Soil Data: BDGSF, ESDAC, WoSIS
- Crop and Production Data: Agreste, EUROSTAT, FAOSTAT
- Data Cleaning, Harmonization, Automatic Updating, and Structured Storage (SQL, Parquet)
3 — Data Analysis & Multi-Source Approaches, Implementation of Advanced Methods to Explore Links Between Bibliographic and Field Data:
- Multi-View Analysis (Multi-View Learning): Integration of Heterogeneous Data Sources.
- Redescription Mining: Discovery of Relationships Between Distinct Descriptor Spaces.
- Exploratory Analysis, Visualization, and Interpretation of Results.
- Evaluation of Model Robustness (Cross-Validation, Bootstrapping).
4 — Incremental Learning & Expert Knowledge Injection, Development of Machine Learning Approaches Enhanced by Human Expertise:
- Incremental/Continuous Learning: Continuous updating of models as new expert data is added.
- Expert Knowledge Injection: Formalization and integration of expert rules (constraints, Bayesian priors, ontologies, grammars, etc.).
- Human-in-the-loop: Interfaces allowing experts to validate, correct, and enrich model outputs.
- Exploration of hybrid neurosymbolic models.
SELECTION CRITERIA
Education
- Master's degree (or equivalent) or PhD in Computer Science, Data Science, Artificial Intelligence, or a related field.
- Training or experience in agronomy, chemistry, or biology, and an awareness of sustainable development issues, are significant assets.
Required Technical Skills
- Programming Languages: Python (mastery required), R preferred.
- AI & LLM: LangChain, LangGraph, CrewAI or equivalent; advanced prompt engineering.
- Machine Learning: scikit-learn, PyTorch or TensorFlow, XGBoost.
- Data Engineering: pandas, SQL, REST APIs, web scraping.
- MLOps: Git, Docker, MLflow or Weights & Biases.
- NLP: named entity recognition, PDF parsing, scientific text processing.
Preferred Skills
- Knowledge of scientific databases: PubMed, arXiv, Scopus, Web of Science.
- Experience with incremental or continuous learning (incremental/continual/lifelong learning).
- Knowledge of chemical formulation, agrochemistry, or agronomy would be an asset.
- Scientific publication(s) in related fields.
Personal Qualities
- Autonomy, scientific rigor, and initiative.
- Ability to work in a multidisciplinary team (agronomists, computer scientists, etc.).
- Intellectual curiosity and a passion for applied research.
- Strong writing skills (technical reports, publications).
Work Environment
The engineer will join a multidisciplinary research team with access to dedicated computing resources and institutional scientific databases.
- Access to HPC/GPU infrastructure.
- Subscriptions to major scientific databases.
- Participation in seminars, workshops, and conferences in the field.
- Opportunities for publishing research findings.
TERMS AND TENURE
This two-year position will be based at the LIBio/LORIA and CEREFIGE Laboratories (F-54000 Nancy) The duration is 18 months.
The target start date for the position is October 1, 2026 with some flexibility on the exact start date (November is accepted).
HOW TO APPLY
Applicants are requested to submit the following materials:
• A cover letter applying for the position (2 pages max)
• Full CV and list of publications
• Academic transcripts (unofficial versions are fine) - Relevant links: GitHub, project portfolio
Deadline for application is July 10th 2026. Applicants will be interviewed by an Ad Hoc Commission by July 23rd 2026
Applications are only accepted through email to:
with the subject line: “Application – AI Research Engineer”:
