Aix-⁠en-⁠Provence Open to opportunities

Baptiste Blouin

An AI engineer’s daily

Single-⁠handed, he ships an AI that gets company data talking

Baptiste Blouin
In Aix-⁠en-⁠Provence.

Aubier makes company data queryable in plain language: hybrid RAG, text-⁠to-⁠SQL, document extraction.

Architecture, back end, front end, infrastructure and day-⁠to-⁠day operations, LLM costs included: all by the same hand.

Get in touch
Download CV
Read the projects

In brief

HistText, the text-⁠analysis platform he designed, deployed to an international community of researchers

See the project

Classifieds

Open the evaluation

Hiring?

  1. Copy the job ad’s text
  2. Paste it here
  3. Verdict in 20 s
read by Claude (Anthropic), then assessed

The interview

Open the assistant

Ask him directly

Background, projects, technical choices: his agent answers, sources included.

Wires · today’s digest

The full AI Digest

About

Data Scientist and AI Engineer with a PhD in Computer Science, combining published ML/NLP research with solid hands-⁠on experience building AI products in production.

  • Engineering

    Full-⁠stack ML, scalable architecture, production deployment

  • Research

    Independent investigation, experimental design, academic publishing

  • Collaboration

    Cross-⁠cultural teamwork, mentoring, international presentations

Experience

  1. - Present

    AI & Full-⁠Stack Engineer · Aubier (independent)

    • End-⁠to-⁠end autonomous design and development of an AI-⁠powered data analytics SaaS platform, from architecture to production
    • Production LLM systems: hybrid RAG (pgvector + BM25), GraphRAG, RAPTOR, text-⁠to-⁠SQL with SSE streaming

  2. -

    Data Scientist & ML Engineer

    Protisvalor · ENP-⁠China Project · Aix-⁠Marseille Université

    • Technical lead and lead developer of the HistText platform, independent project management; helped secure an ERC Proof of Concept grant to bring it to market
    • Explored and solved complex ML challenges in large-⁠scale historical text processing

  3. -

    PhD Candidate & ML Scientist

    Protisvalor · LIS / IrAsia Laboratory · Aix-⁠Marseille Université

    • Independent research on European ERC Advanced grant, self-⁠directed learning
    • Published 4 peer-⁠reviewed papers during the PhD (7 in total), technical writing and communication

Key Projects

Independent project · in production

Aubier, Data & AI SaaS Platform

Data-⁠analysis SaaS platform powered by AI, multi-⁠tenant and in production, designed and built end-⁠to-⁠end single-⁠handedly (backend, frontend, infrastructure), covered by over 6,600 automated tests.

–

  • Hybrid RAG
  • text-⁠to-⁠SQL
  • GraphRAG
  • RAPTOR
  • LLM cost control
  • LLM security
  • Observability
  • multi-⁠tenant
  • FastAPI
  • React
  • TypeScript
  • pgvector
  • litellm
  • Langfuse
  • Celery
  • Docker
  • Stripe

Product site In production

–

baptisteblouin.fr, Portfolio & AI Agent

Production AI Agent · Public MCP Server

This site, designed, built and operated single-⁠handedly, is a showcase running in production.

Product site

–

Private Apps, Agents & Authenticated MCP Servers

4 applications in production · 2 private MCP servers

Four private applications built and operated single-⁠handedly (FastAPI, HTMX, over 3,000 tests): a stock screener, a business application encrypted at rest, a monitor for the site’s agent and an operations dashboard.

In production

 - 

HistText

Large-⁠Scale Text Analysis Platform

Full-⁠stack platform for large-⁠scale analysis of historical Chinese texts (over 15 million documents, 20 billion words).

Live demo GitHub JDMDH 2024

 - 

EventExtractionPapers

Open Source Community Resource

Curated selection of NLP resources dedicated to event extraction: papers, datasets, models and code. A reference widely used by the international NLP research community, with 580 GitHub stars as of 3 September 2026. Maintained from 2019 to 2024.

  • NLP
  • Event extraction
  • Open source

GitHub 580 stars (09/2026)

Applied AI Rig

Open Source Tool · Traceability of AI Decisions

Command-⁠line tool that installs, inside a repository, the decision and evidence record of a project that ships AI: why this model, which evaluation justifies a threshold, which cost limit, which fallback.

GitHub

–

Tech Brief

Automated Publishing · Open Source

Public mirror of the site’s tech watch, published as bilingual Markdown on GitHub.

GitHub

ENP-⁠Corpus Creator

OCR & Annotation Pipeline

Tool suite transforming scanned images of historical newspapers into semantically enriched corpora.

Article

 - 

Chinese NER Dataset & ML Pipeline

Data Engineering & ML

The largest manually annotated NER dataset of historical Chinese newspapers from the transitional period (1872-⁠1949), per the LREC-⁠COLING paper: 196 Shenbao documents, 6,491 entities (persons, places, organizations, dates) over 121,971 characters, with entity linking and coreference.

LREC-⁠COLING 2024

Chinese Language Model & Tokenization

ML Model Development

Fine-⁠tuning of segmentation models for transitional Chinese, in collaboration with Academia Sinica.

NLP4DH 2023

OCR Simulation & ML Robustness

Experimental Research

Data-⁠augmentation framework simulating the OCR artifacts (character substitutions, deletions, insertions) observed in digitized historical documents.

TALN 2022

Transfer Learning & Adaptation to the Historical Domain

ML Development & Optimization

Study of model transfer methods for NER on historical documents (out-⁠of-⁠domain genres, noisy OCR, older variants of the language), depending on the effort invested: describing the target data, annotating a few texts or matching pre-⁠training resources.

NLP4DH 2021

Skills

Languages

  • Python
  • Rust
  • JavaScript / TypeScript
  • Java
  • C++
  • R
  • SQL

LLM & Generative AI

  • litellm
  • LangChain
  • LangGraph
  • Langfuse
  • Prompt Engineering
  • Structured Output
  • Multi-⁠provider Routing & Fallback
  • RAG
  • GraphRAG
  • RAPTOR
  • Hybrid Search (BM25 + Vector)

Full-⁠Stack Development

  • React
  • Vue.js
  • TailwindCSS
  • D3.js
  • Shiny
  • Node.js
  • FastAPI
  • Flask
  • SQLAlchemy
  • Pydantic
  • Async Python
  • REST API
  • WebSockets
  • Auth / RBAC
  • OAuth

Infrastructure & DevOps

  • PostgreSQL
  • MySQL
  • pgvector
  • Elasticsearch
  • Apache Solr
  • Redis
  • Celery
  • Alembic
  • Docker
  • Nginx
  • S3 / MinIO
  • Prometheus / Grafana
  • OpenTelemetry
  • GCP

NLP & Document Processing

  • spaCy
  • NLTK
  • Multilingual NER
  • Event Extraction
  • Transfer Learning
  • Domain Adaptation
  • Word Segmentation
  • Data Annotation
  • OCR (Docling, Google Cloud Vision)
  • Historical Text Processing

Research & Communication

  • Scientific Publications
  • Scientific Methodology
  • Experimental Design
  • Technical Writing & Publishing
  • International Presentations
  • Literature Review
  • Mentoring
  • Cross-⁠cultural Collaboration
  • LaTeX
  • ADR (Architecture Decision Records)

ML & data science

  • PyTorch
  • TensorFlow
  • Hugging Face
  • scikit-⁠learn
  • FLAML / AutoML
  • Sentence Transformers
  • NetworkX
  • UMAP
  • Statistical Analysis
  • pandas
  • DuckDB
  • ONNX Runtime

Publications

  1. HistText: An Application for Leveraging Large-⁠Scale Historical Textbases

    Baptiste Blouin, C. Armand, C. Henriot

    JDMDH 2024, Journal of Data Mining & Digital Humanities

    PaperHALGitHub

  2. A Dataset for Named Entity Recognition and Entity Linking in Chinese Historical Newspapers

    Baptiste Blouin, C. Armand, C. Henriot

    LREC-⁠COLING 2024, Turin, Italy

    ACL Anthology

  3. Unlocking Transitional Chinese: Word Segmentation in Modern Historical Texts

    Baptiste Blouin, H.-H. Huang, C. Henriot, C. Armand

    NLP4DH 2023, Tokyo, Japan

    ACL Anthology

  4. Simulation d’erreurs d’OCR dans les systèmes de TAL pour le traitement de données anachroniques [Simulating OCR errors in NLP systems for processing anachronistic data]

    Baptiste Blouin, B. Favre, J. Auguste

    TALN 2022 (JEP-⁠TALN-⁠RECITAL), Avignon, France

    ACL Anthology

  5. Transferring Modern Named Entity Recognition to the Historical Domain: How to Take the Step?

    Baptiste Blouin, B. Favre, J. Auguste, C. Henriot

    NLP4DH 2021, Silchar, India

    ACL Anthology

  6. Creating Biographical Networks from Chinese and English Wikipedia

    Baptiste Blouin, N. van den Bosch, P. Magistry

    Journal of Historical Network Research, Vol. 5, No. 1

    JHNRHAL

  7. Contextual Characters with Segmentation Representation for Named Entity Recognition in Chinese

    Baptiste Blouin, P. Magistry

    PACLIC 34, Hanoi, Vietnam

    ACL Anthology

Education

  •  - 

    PhD in Computer Science (NLP / Machine Learning)

    Aix-⁠Marseille Université

    Thesis: Event extraction from facsimiles of historical documents for history studies

    Advisor: Prof. Benoît Favre, ERC Advanced Grant ENP-⁠China (no. 788476)

    Read the thesis

  •  - 

    MSc Computer Science (AI/ML)

    Aix-⁠Marseille Université

  • BSc Computer Science

    Aix-⁠Marseille Université

AI Digest

Sourced AI / ML summaries, generated automatically twice a day from a curated set of RSS feeds, with a weekly recap on Sunday evening.

Loading the digest…

Contact

Feel free to reach out to discuss opportunities, collaborations, or just to chat.

Availability Open to new opportunities

An offer, a question?

The message lands in my inbox. Leave an address if you would like a reply.

Download my CV

Contact details

Letters to the editor

The message lands in my inbox. Leave an address if you would like a reply.

Leave the address empty to send me a note anonymously. Without it, I simply will not be able to reply.

Your address is only used to reply to you. Privacy

Correspondence

I will send you the version you choose as an attachment, with a short introduction. No subscription is created.

Your address is only used for this one email. Privacy

Subscriptions

One email a week, on Sunday evening: the sourced summary of the week in AI and in papers. Nothing else, and one-⁠click unsubscribe.

Your address is only used for this digest. Privacy