Sébastien AucanteBackend Engineer and AI Solutions
Portrait of Sébastien Aucante

Sébastien Aucante

Backend Engineer and AI Solutions

Skills

  • RAG
  • Retrieval
  • LLM integration
  • Evaluation
  • AI products
  • Full-Stack Development
  • API & Backend
  • Software Architecture
  • RBAC
  • JWT
  • SOLID
  • Python
  • PHP
  • Node.js
  • TypeScript
  • FastAPI
  • LangChain
  • Qdrant
  • Pydantic
  • Langfuse

AI Engineering & Integration

Engineering reliable AI solutions for real-world products.

I design, build and integrate AI systems — from retrieval and RAG architectures to production-ready AI services.

View prototypesDiscuss a project

RAG / Retrieval / LLM Integration / Evaluation / AI Products

Positioning

From AI prototypes to services integrated into real information systems.

I design and integrate internal AI solutions: business assistants, document search, RAG systems, automation and LLM-powered services. The objective is to turn a business requirement into an operational system while considering reliability, security, data governance, traceability and data sovereignty from the architecture stage.

My full-stack development background allows me to work beyond the model itself: APIs, backend services, interfaces, data, integration with existing systems, testing and production delivery.

Personal prototypes

Small systems built to investigate difficult engineering questions.

Repositories available when the prototypes are published.

  1. 01Comparative prototype

    From naive RAG to reliable retrieval

    INFOPAY-AI and AEROSPEC-AI compare a rapid AI-assisted prototype with an evaluated RAG engineering approach.

    RAG · Chroma · LangChain · FastAPI · React · TypeScript · Python · 2026

    View prototype

    INFOPAY-AI · Baseline

    INFOPAY-AI interface

    AEROSPEC-AI · Engineered

    AEROSPEC-AI interface

  2. 02Independent prototype

    SecureBank-AI — Securing a RAG platform with RBAC and JWT

    A hybrid RAG platform combining JWT authentication and RBAC access control with a reproducible comparison of four retrieval strategies.

    RAG · RBAC · JWT · BM25 · Qdrant · RRF · FastAPI · React · TypeScript · 2026

    View prototype

    SecureBank-AI · Engineered

    SecureBank-AI interface
  1. 01

    RAG & Retrieval

    Designing reliable knowledge retrieval architectures.

  2. 02

    AI Integration

    Integrating LLM capabilities into existing products and workflows.

  3. 03

    AI Engineering

    Evaluation, reliability and production-oriented AI systems.

RAG Architecture

Engineering the retrieval layer behind reliable AI products.

A RAG system is more than connecting documents to an LLM. Reliability depends on how information is parsed, chunked, retrieved, ranked, assembled into context and evaluated.

Indexing

  1. Documents
  2. Parsing
  3. Chunking
  4. Embeddings
  5. Vector Store
Orchestration layer — LangChain

Retrieval & generation

  1. User Query
  2. Query Embedding
  3. Retrieval
  4. Reranking
  5. Context Construction
  6. LLM
  7. Grounded Answer + Citations
Vector StoreRetrieval

Expertise

Focused technical capabilities for dependable AI delivery.

01

AI Engineering

  • RAG & retrieval
  • LLM integration
  • Context engineering
  • Embeddings & vector search
  • Evaluation & reliability
  • AI workflows & automation
  • Observability & traceability

02

Software Engineering

Languages & application development

  • TypeScript / JavaScript
  • PHP
  • Node.js
  • Python
  • React / Next.js

Architecture & backend

  • API & backend engineering
  • Software & system architecture
  • SOLID principles
  • Design patterns

Delivery & integration

  • Data & systems integration
  • Testing & quality
  • CI/CD & production

Contact

Have an AI problem worth solving?

Let’s discuss the product, the data and the engineering behind it.