Explore my background

Skills connected to real work.

I connect my product-building experience with the AI Engineer programme. Choose a domain, discover the projects, then explore their results and limitations.

Complementary learning experiences

Before and outside OpenClassrooms

Independent learning, earlier training and professional experience. These categories describe skill origins, rather than a strict boundary in time.

  • Bubble · no-code · API · Postman
  • Python · TypeScript · React · Next.js · NestJS
  • Expo · Docker · PostgreSQL
  • Neon · Supabase · Convex
  • Design, scraping and integrations

Developed through the AI Engineer programme

Educational projects providing evidence of practice, recorded results and evaluation feedback.

  • Supervised ML, validation and explainability
  • RAG · embeddings · LangGraph
  • MLflow · CI/CD · monitoring · ONNX
  • Vision, semi-supervision and reinforcement learning
  • ETL multimodal · SFT · LoRA · DPO

Prior skills reflect my stated background and professional work. The OC report distinguishes inspected code, recorded results and assessor feedback, without invented proficiency percentages.

What I learned through project work

From scoping to operations, each domain opens the projects and skills supporting it.

Open the interactive skills map ↗

What I want to strengthen

Authentication and authorisation

The P5 authentication competency was not validated in OC feedback, despite the overall project status. Professional experience is a different context and does not replace that assessment.

Independent evaluation and RAG faithfulness

Strengthen the separation between selection and final testing, question coverage and error analysis. A response or score alone does not establish reliability.

Operations and autonomy

Connect performance with versions and environments, and document costs and maintenance. My autonomy by skill and language levels remain to be specified in my self-assessment.

The next personal project

Being scoped — POC planned

A report-drafting assistant using entirely synthetic patient records: transcription, retrieval from patient history, sourced drafts and human review. Its report, measurements and deployment will be added once completed.

The learning report

An overview of learning, technical decisions and areas for development through completed projects.

Read the learning report ↗