AI training that sticks
Live corporate training on Claude and AI workflows, delivered on camera in English or French. Every session ends with something the team runs the next day.
Anthropic-certified Claude instructor and selected Claude partner. Four years building AI education for 8,000+ learners across 15+ countries, including curriculum co-designed with Kazakhstan's Ministry of Education and government-backed programs with France's Ministry of National Education. I encode teaching craft into AI-powered systems instead of guarding it.
What I do
Whether you need live training, a learning program, or a content system with a quality bar encoded in it.
Live corporate training on Claude and AI workflows, delivered on camera in English or French. Every session ends with something the team runs the next day.
Learning programs from single workshops to multi-country curricula. Project-based, participatory, measured. Built for 8,000+ learners at one of the world's largest coding school networks.
Production pipelines where Claude drafts, validates, and maintains content under human-defined quality bars. I design where the human line sits and how it moves as models improve.
Selected work
IT Content Development Lead, 01Edu (global project-based coding network)
Leading content development and delivery operations at 01Edu, a global project-based coding network. I built the AI and data engineering curriculum tracks alongside the core software program, and I own the pipeline end to end: planning, team coordination, technical review, implementation, and launch. Programs delivered under government contracts in France, Kazakhstan, and Morocco, where learners work without traditional lectures.
Delivered government-backed coding and AI programs across three countries, reaching 8,000+ learners through a peer-to-peer model with zero traditional lecture hours.
Founder, EUNOVAI, an AI enablement consultancy
I founded EUNOVAI to teach Claude to the people who do not write code: sales, operations, marketing, and support teams. The work goes past prompt tips. Each engagement maps a team's real day-to-day tasks, then turns them into Claude workflows the team owns, with a playbook and review standards they keep after I leave.
Non-technical teams leave with working workflows they run themselves, not slides, and a written standard for where the human review line sits.
Internal tooling and workflow design for a global curriculum
Built AI-assisted validation and deployment workflows for a global curriculum. Designed the review pipeline so Claude checks content against defined quality criteria before any human reviewer sees it, surfacing only genuine edge cases for human judgment.
Removed hours of manual review weekly and made quality consistent regardless of reviewer, without sacrificing editorial control.
Live training engagement, IT consultancy
Designed and delivered live Claude training for an IT consultancy team, including a Claude Cowork workflow playbook tailored to their specific stack and daily tasks. The session moved from conceptual overview to hands-on practice with real deliverables.
"The training was very good... both the examples and the style of the training were excellent. I am sure we will work together again."
AI-scaled creative production pipeline
Built a pipeline that extracts a brand's design language and generates derived websites, presentations, animations, and mobile apps within it. Claude Code reads brand assets, infers tokens and patterns, then produces on-brand output for each target format under a human creative review gate.
Proof of AI-scaled creative production. This site was designed and shipped with that workflow.
Event-driven lead qualification and outreach system, commercial real estate
Designed and built a staged pipeline for a commercial real estate team. Incoming leads are normalized into a structured record, then scored against the team's own qualification criteria before anything is written. Claude runs the generation stage inside a constrained prompt system: it drafts from the record and a voice specification rather than free-form, so every message is grounded in facts the pipeline actually holds. Drafts pass a validation stage against defined quality and accuracy checks, and a failure routes back for rewrite instead of forward to a queue. Follow-up sequencing runs on a scheduler with rate limits, and a human approval gate sits in front of every send.
Manual prospecting went from hours to about 20 minutes a day. The team reviews drafts instead of writing them, and the quality bar lives in the pipeline rather than in each rep's head.
Course creator, Eduonix Learning Solutions
Created the Claude AI Professional E-Degree for Eduonix Learning Solutions, a four-course program covering prompt engineering, agentic coding, and workflow automation, built to take learners from basic chatbot use to running Claude as an automated workflow engine. The program includes 100+ lectures, 8+ hours of video, and five practical projects with a completion certificate.
Live on Eduonix's global catalog, a platform with 1,177,456 students and 82,771 reviews worldwide.
Engineering background
I train engineering teams on Claude because I came out of one. The systems background is why I can say where AI belongs in a pipeline, and where it does not.
Worked inside an existing production architecture rather than a greenfield one: designed changes against it, scaled an on-premise system, wrote unit tests, and debugged algorithms. The habit it left is the useful one, which is reading a system you did not write before changing it.
Built web applications front to back with Vue.js and Laravel: interface, application layer, and database, shipped remotely against real requirements.
Systems programming in C, memory management, and low-level problem solving, learned in a school with no teachers and no lectures. Every project defended in front of peers. This is also where the thesis behind everything below comes from.
Engineering and business fundamentals program run with Microsoft teams for undergraduate and masters students, mentored under the global talent acquisition organization.
$ BSc Computer Science · C and systems · Vue.js, Laravel · Claude Code
How I think
I learned to code at a school with no teachers. 1337, a 42 Network school: no lectures, just projects and peers, an environment designed so learning happens without an instructor in the room. That formed my core belief.
Teaching scales when you design the system, not when you multiply the teachers. AI completes that idea. When every learner has an adaptive tutor, education stops being content delivery and becomes environment design.
That is what I build: learning systems with a quality bar encoded in them, where AI does the scaling and humans own the taste. The goal is not to automate education. It is to make good education reach everyone.
Writing and teaching
Long-form thinking on how AI changes what it means to teach and learn at scale.
A structured approach to building technical learning plans, published for the Microsoft educator community.
A recorded session showing how I structure live AI training for real teams.
A short walkthrough of a content validation and generation pipeline I built with Claude.
About
I'm Oumaima. I work across European and US hours, in English and French. Software engineering background from 1337 / 42 Network, BSc Computer Science.
I founded and run EUNOVAI, an AI enablement consultancy that teaches non-technical teams how to work with Claude, and I teach for the joy of the moment when something clicks.
Hiring for an education, content, or AI enablement role, or want training for your team? Email me.