Introduction

I'm Leo Weiss, an AI engineer interested in what makes intelligent systems actually work.

My work spans AI system architecture, model evaluation, and production engineering, from designing LLM-powered applications to understanding their behavior, limitations, and performance in real-world environments.

My background combines engineering and finance, with experience at Air France, Edmond de Rothschild Asset Management, and Boston Consulting Group. I hold an MSc in Data for Finance from Mines Paris – PSL and am pursuing a Specialized Master's in Data & AI at Télécom Paris.

Outside of work, I build tools to explore how models represent information, make decisions, and behave.

Languages

  • Frenchnative
  • EnglishC1

Skills

  • Python
  • SQL
  • TypeScript
  • RAG
  • Embeddings & Vector Search
  • LLM Evals
  • LLMOps
  • LangChain
  • MCP
  • Databricks
  • PySpark
  • Azure
  • Docker
  • Three.js

Experience

Boston Consulting Group logo

EMESA Analytics Engineer — Apprenticeship

–

Boston Consulting GroupParis, France

Building data pipelines and AI-driven analytics tooling for brand and campaign performance work.

  • Design and automate data pipelines and AI-driven analytics tools that structure marketing data and sharpen KPI tracking
  • Apply Python, SQL and dashboarding to brand and campaign analysis, supporting strategic recommendations
StackPythonSQLData PipelinesAnalytics
Edmond de Rothschild Asset Management logo

Edmond de Rothschild Asset Management

–

Paris, France

AI Engineer — Apprenticeship

–

Building and shipping LLM applications for funds managers.

  • Designed and deployed LLM applications — RAG pipelines, retrieval systems and domain chatbots — giving investment teams direct access to financial and ESG knowledge
  • Built end-to-end AI workflows: ingestion, chunking and vectorization, model orchestration, evaluation, and monitoring in production
  • Designed scalable AI architectures on Databricks and Azure integrating cloud infrastructure, internal data systems and APIs behind decision-support tools
  • Implemented LLMOps practice — prompt and model versioning, experiment tracking, regression evaluation, and deployment pipelines
StackPythonRAGLLMOpsDatabricksAzure

Quant Developer — Apprenticeship

–

Developed quantitative models for ex-ante performance, scenario analysis and risk forecasting supporting portfolio management decisions.

  • Built fixed-income analytics — sensitivities, bond behaviour, contribution and attribution analysis — and automated portfolio monitoring workflows
  • Designed and maintained the data pipelines (Power Query / SQL Server) feeding Power BI dashboards, automating ingestion, cleaning and recurring reporting
StackPythonSQLPower BIVBAExcel
Air France logo

Data Analyst — Apprenticeship

–

Air FranceParis, France

Delivered a cost-reduction programme on heavy maintenance checks for the B777 fleet, working from raw maintenance and operational data.

  • Built and automated an end-to-end analytics pipeline with data quality checks, traceability and reproducibility
  • Designed Power BI dashboards on Microsoft Cloud and SAP data to surface cost drivers and track the impact of operational actions
StackPythonSQLPower BIPower QuerySAPAzure

Education

Télécom Paris logo

Specialized Master’s in Data & Artificial Intelligence

–

Télécom ParisParis, France

Databases, cloud computing, statistics and statistical learning, advanced ML, GenAI and deployment, MLOps, web scraping and NLP, and the legal and regulatory layer.

Mines Paris – PSL logo

MSc Data for Finance

–

Mines Paris – PSLParis, France

Joint degree with Albert School. Statistics and econometrics, machine learning, market finance, asset valuation, algorithmic trading, quantitative development, time series, data engineering and MLOps.

ESME logo

Bachelor of Engineering Science

–

ESMEBordeaux, France

Physics, mathematics, statistics and probability, programming, and aeronautics engineering.

Vilnius Gediminas Technical University logo

Exchange semester

–

Vilnius Gediminas Technical UniversityVilnius, Lithuania

Aerodynamics, aircraft structures and engines, and human factors.

Extracurricular

OpenAI Forum logo

Community Leader

forum.openai.com

OpenAI Forum

The Forum is an invitation-only community of practitioners and researchers convened by OpenAI, with member-led chapters around the world.

  • Part of the Paris chapter — expert round tables and informal meetups

Member

Until

Albert Quant Trading Club (AQTC)

Member

Until

Albert Junior Consulting (AJC)

InterestsEntrepreneurshipTechnologyPhysicsSwimmingPadel

Selected work

Lexiscope — screenshot

Lexiscope

A 3D semantic space explorer — UMAP-projected word embeddings you can fly through.

Three.js · UMAP · Vanilla JS

lexiscope.netlify.app
Unallocated — screenshot

Unallocated

What tokenizer vocabularies were spent on, and what they weren’t — a measurement study of 204 languages across 8 tokenizers.

D3.js · Vite · FLORES-200

unallocated.netlify.app

More on GitHub →

Writing

Contact Me

leo.weiss33@gmail.comDownload CV

Leo Weiss — Paris, 2026