As part of a project to optimize our wastewater treatment plant (WWTP) processes, we are seeking a senior AI consultant to support the implementation and operation of AI tools during a mission that will take place in 3 phases.

Phase 1 – Report (20 working days, 1 day/week)

Writing a report defining the requirements necessary for the implementation of a predictive system and decision-support tools for the plant's operation.

This report should include:

• An overview of the different types of AI (generative AI, machine learning, predictive models, neural networks, etc.) as well as their areas of application, advantages, and limitations;
• Identification of the solutions best suited to the challenges faced on our installations and assessment of their technical and economic relevance;
• A schedule, a budget, and the resources required for different implementation solutions (internally or fully outsourced).

Phase 2 – Training and coaching (conditional) for the implementation of the chosen solution (220 working days, 0.5 day/2 weeks)

The training should aim to:

• Acquire practical knowledge of the tools available on the market, their actual capabilities, and the conditions required for their implementation;
• Master the methods of preparation, structuring, and exploitation of industrial data necessary for the development of reliable predictive models;
• Know how to formalize a business need in the form of a requirements specification or a CSC, allowing effective consultation of specialized suppliers or integrators;
• Be able to configure, parameterize, and interact effectively with AI tools to obtain relevant and usable results;
• Analyze, interpret, and validate the results produced by AI models, while understanding their limitations and uncertainties;
• Improve model performance by adjusting parameters, input data, and methods used.

Phase 3 – Generic consultancy (conditional, 20 working days over the total duration of the project)

Benefit from support on concrete cases related to the operation of the WWTP, particularly for the implementation of predictive models applied to treatment processes, energy performance, anticipation of malfunctions, or optimization of operations.

Hybrid work arrangement: 3d/5 on site