Freelancer profile translated to English.
Back to original languageAbout Damien
AI Engineer — Gen AI Developer / Multi-agent Systems
I design and deploy end-to-end generative AI systems: orchestrated multi-agent architectures, specialized RAG pipelines, persistent memory, and agentic workflows in production. I handle everything from architectural design to cloud deployment, focusing on reliability, cost control, and industrialization.
Engagement types I am looking for
- Gen AI Systems Architecture : multi-agent design, model routing, context/memory layers (episodic, semantic), autonomy gradient
- Vertical RAG Pipelines : specialized corpus ingestion and structuring, vector databases, chunking and retrieval strategies, relevance evaluation
- Prompt Engineering : structured system prompts, prompt chaining, function calling, constraints, and guardrails
- Agentic Workflows : agent orchestration, tool use, synchronous and asynchronous processing, event-driven
Technical Stack & Domains
- Models : Claude (Haiku, Sonnet, Opus), GPT, open-source models; multi-model routing based on cost/quality
- Frameworks : Claude Agent SDK, managed agent services, MCP
- Cloud : AWS, GCP, Azure — managed agent and knowledge base services, serverless functions, NoSQL databases, event bus
- Reference Architectures : multi-agent, RAG, persistent memory, function calling
My Contribution
- Systems Thinking : I design complete architectures, not just isolated prompts or API calls
- Industrialization : my systems are built for production — reliability, controlled costs, managed latency, observability
- Verticality : I structure domains of expertise into knowledge layers usable by agents
If you are building a Gen AI product and looking for someone to handle the end-to-end architecture, I would be delighted to discuss it.
French
Native or bilingual
English
Fluent
Can work on-site
Paris (up to 50km)
Experience
- LEXGUILDAI Engineer — Gen AI, Multi-agent & Legal Knowledge Base SystemsTECHApril 2026 - Today (4 months)Paris, France
LEXGUILD — Specialized Multi-agent Architecture for Legal Education (SaaS / EdTech / LegalTech)
Context & Objective
Design, architecture, and deployment of a multi-agent system for French law students (L1, L2, L3).LexGuild acts as a cognitive orchestration layer: a multi-agent system that understands the student's progress, selects the right exercise at the right time, mobilizes the correct legal texts and case law via specialized RAG, and delivers measurable progress towards academic success.Detailed Missions
Architecture & System Design- Design of an orchestrated multi-agent architecture on AWS Bedrock (relation, context, knowledge, orchestration, trust, workflow, result layers)
- Design of a model router to optimize the cost/quality ratio per pedagogical task
- Design of the pedagogical knowledge graph: student profile, episodic memory, multi-axis methodological calibration, longitudinal progress curve
AI Production & Automation- Design of specialized multi-agent workflows (tutorials, methodology, essay grading, revision, mock exams)
- Structuring of system prompts per agent and per pedagogical modality
- Industrialization of Lambda Action Groups (StudentProfiler, MemoryExtractor, ProgressTracker)
- Implementation of the legal RAG pipeline: Bedrock Knowledge Bases indexing codes, case law, and doctrine with domain-specific granularity
Cloud Infrastructure & Data- Full-AWS architecture: Bedrock (models + Knowledge Bases + Agents), Lambda (orchestration), DynamoDB (profiles, episodic memory, sessions), EventBridge (asynchronous processing)
- DynamoDB table design
- Design of session lifecycle Lambdas
- Cost tracking per session
- D2MEQuantitative Systems Engineer — Rust Trading Bot & Autonomous Multi-agent DeskTECHOctober 2025 - April 2026 (6 months)Paris, France
BITCOIN TRADING BOT — Autonomous Quantitative Architecture on Predictive Markets (Crypto / HFT / Prediction Markets)
Context & ObjectiveDesign, architecture, and deployment of an autonomous quantitative trading system operating 24/7 on 5/15-minute Bitcoin predictive markets, complemented by a multi-agent Claude desk that continuously pilots, optimizes, and improves the bot. Strategic objective: disciplined exploitation of an edge through 384 daily trades. The system is a dual architecture: a deterministic Rust execution engine coupled with a multi-agent cognitive layer ensuring adaptation to market changes.Detailed MissionsArchitecture & System Design- Design of a two-strata architecture: deterministic Rust hot-path and deliberative multi-agent desk
- Design of a pricing model suitable for binary options
- Design of the multi-signal aggregation pipeline: decorrelated alpha signals, 3-state HMM market regime model, inter-signal correlation tracker
AI Production & Automation- Design of six specialized Claude agents with controlled writing authority over bot configurations
- Structuring of system prompts per agent with authority matrices, absolute ranges, maximum deltas, and cooldowns
- Industrialization of the internal control server
- Implementation of the orchestration pipeline
Infrastructure- Hybrid architecture: native Rust execution, Claude Managed Agents, Redis Streams, PostgreSQL
- Design of the four stacked configuration layers and the four cascading guards
- Design of fractional Kelly sizing with capital percentage caps, four-phase scaling strategy, maker-dominant architecture
- Prometheus monitoring, real-time dashboard, Telegram alerting, API cost tracking per agent
- PRO DU PROMPTPrompt & Product Engineer — AI, Expert Content & Data-Driven SystemsEDUCATION AND E-LEARNINGSeptember 2023 - September 2025 (2 years)Paris, France
Project Name
PRO DU PROMPT / LAW — AI-driven series of legal books (Legal / EdTech)Context & Objective
Design and management of a complete system for the production, editing, distribution, and acquisition of an AI-assisted legal book series, within a product and catalog logic (3 series of 10 books each, 500 pages, covering the fundamental subjects of French law degrees L1, L2, & L3).Detailed Missions
Production & Automation- Design of content generation workflows via AI (Claude, ChatGPT)
- Structuring of complex prompts to produce optimized and adapted university-level legal prompt catalogs (courses, methods, exercises)
- Industrialization of production (30 products delivered - 18,000 legal prompts)
Product & Distribution- Semi-automated text/cover design (Canva, Keynote, AI image)
- Structuring of product catalogs (Amazon KDP)
- Implementation of optimized product pages (SEO, A+ Content)
- Creation of Amazon product marketing (product page, series pages, A+ content)
- Generation of promotional videos (AI voice, AI music, editing)
Data & Acquisition- Creation and management of Amazon Ads campaigns
- Sponsored Products Campaigns: 10 test campaigns per product (4 automatic types, 5 manual keyword types, 3 manual product types)
- Sponsored Brands Campaigns: 15 test campaigns per series (5 Product Collection types, 5 Store Spotlight types, 5 Video Ads types)
- Performance analysis (Impressions, CPC, conversions, ACOS)
#Stack
AI: Claude, ChatGPT (Dall-E)Design & Content: Pages, Keynote, CanvaData & Positioning: Helium 10 (Magnet, Cerebro, Tracker)Distribution & Ads: Amazon, Amazon KDP (paperback/ebook), Amazon Ads (SP, SB)Audio/Music/Video: ElevenLabs, Suno, CapCut
Recommendations
Be the first to recommend Damien
Help this freelancer shine by sharing your experience working together.
These freelancer profiles also match your criteria
Agatha Frydrych
Backend Java Software Engineer
4.7
(3)
2
Baptiste Duhen
Fullstack developer
4.6
(4)
5
Amed Hamou
Senior Lead Developer
4
(2)
7
Audrey Champion
Web developer
4.3
(3)
4
Education
- Master 2 Financial EconomicsParis I Panthéon-Sorbonne2018Master Thesis - Méthode de contrôle synthétique pour des études de cas comparatives : estimation des effets de l'union économique et monétaire sur le produit intérieur français entre 1993 et 2015
- Bachelor's degree in Economics and ManagementFACO2016Mémoire de licence - La neutralisation de l'Allemagne wilhelmienne entre 1919-1924 : les conséquences économiques, financières et monétaires du traité de Versailles.
Certifications
- AWS Solutions Architect AssociateAWS2024
- AWS Certified Cloud PractitionerAWS2023