Florian Lecoeuche

AI transformation mission with RAZEL-BEC, a major player in public works with €1.1 billion in revenue.

🔹 Context:

RAZEL-BEC aims to integrate AI into the preparation of bids for public works contracts, for demanding projects where quality, confidentiality, traceability, internal data, and human expertise remain critical. The goal is not to automate the bidding process, but to build a reliable, adoptable, and reusable AI methodology.

Key Responsibilities:

🔸 Framing the AI approach: objectives, scope, risks, usage limits, available data, success criteria, and adoption conditions.

🔸 Change management support for business users who are new to or have limited experience with AI.

🔸 Identification and prioritization of high-value operational AI use cases.

🔸 Implementation of an AI governance framework: confidentiality, human validation, output quality, traceability, risk management, and no substitution for business expertise.

🔸 Structuring reusable AI workflows to assist with document review, requirements extraction, template preparation, initial draft writing, and critical review of deliverables.

🔸 Definition of evaluation criteria: relevance, reliability, completeness, absence of fabrication, expert correction time, user confidence, and actual gain.

🔸 Building a reusable foundation: prompts, templates, checklists, review methods, and best practices.

🔸 AI industrialization roadmap.

Expected Results:

🔸 Structured AI methodology, tested on real cases, and adapted to business constraints.

🔸 Skill development for teams on mastered AI uses.

🔸 Reduction in time spent on document review, structuring, and preparation.

🔸 Improvement in the quality of initial drafts produced with AI.

🔸 Trust framework to move from AI experimentation to sustainable, governed, and industrializable uses.

Transformation mission to support an IT team in adopting artificial intelligence, structuring their work practices, and improving their operational performance.

🔷 Context: A technical team facing challenges of modernization, reliability, technical debt, and adoption of new AI uses within an existing application environment.

Key Responsibilities:

🔸 Diagnosis of AI adoption barriers and change management for technical profiles;

🔸 Acculturation of the team to concrete AI uses: debugging, code comprehension, technical analysis, formalization of requests, tickets, and documentation;

🔸 Transformation of initially cautious attitudes into autonomous and regular usage, by securing through practice;

🔸 Structuring of framing practices: clarifying needs, formulating objectives, improving tickets, and strengthening validation criteria;

🔸 Awareness of agile and Scrum fundamentals to improve communication between business needs, functional design, and development;

🔸 Identification of high-value AI use cases, without substituting the tool for human understanding;

🔸 Pragmatic trade-offs between ambition, organizational maturity, and technical reality;

🔸 Contribution to improving operational reliability through prioritization of immediately actionable technical tasks.

Results:

🔸 Effective adoption of AI by all team members, with autonomous reuse on concrete cases;

🔸 Time savings identified on debugging, code exploration, and technical understanding tasks;

🔸 Improvement in the level of structuring for requests, tickets, and specifications;

🔸 Initial transformation dynamic established around AI, delivery practices, and reliability;

🔸 Implementation of concrete technical reliability actions on priority backend services.

Taking over a critical in-house ERP, custom-developed and maintained by the IT team, within an agri-food company with €361 million in revenue. Central system for commerce, logistics, billing, shipping, and traceability.

🔷 Context:

→ ERP used by almost all departments, impacting daily operations in France and internationally.

→ Aging IT system: obsolete Oracle 9i database and over 35 legacy satellite applications.

→ High dependence on internal knowledge: loss of historical knowledge, technical debt, slowness, bugs, and support consuming over 60% of activity.

→ Roadmap disconnected from execution capacity, shifting priorities, and lack of visibility.

Key Responsibilities:

🔸 Direct management of an 8-person team dedicated to the development, maintenance, and support of the ERP/satellite applications.

🔸 Taking over the roadmap, backlog, prioritization, and coordination between business departments, support, technical team, and CIO.

🔸 Clarification of roles, regular one-on-one meetings, and preparation of the 2026-2027 IT/ERP budget.

🔸 Structuring business requests, objectifying business priorities, and reducing parallel initiatives.

🔸 Trade-offs between maintenance, refactoring, migration, technical debt, and business value.

🔸 Reframing and relaunching the Oracle 9i to PostgreSQL 17 migration.

🔸 Controlled AI usage to accelerate legacy reverse documentation and structure specifications.

Results:

🔸 Roadmap and backlog restructured around a realistic and business value-oriented approach.

🔸 Team remobilized through proximity management and clarification of responsibilities.

🔸 Implementation of a high-impact supply chain ERP evolution, generating estimated savings of €12 million/year.

🔸 Critical migration reframed and relaunched with an external database architect.

🔸 AI-powered reverse documentation of the legacy system, covering over 35 critical and satellite applications.