Freelancer profile translated to English.

Description

I am Dionysius, an AI engineer based in Lille, specializing in the design and deployment of custom artificial intelligence systems for businesses. My expertise covers the entire technical spectrum: from no-code solutions to advanced architectures.

What I build:

AI multi-agent systems capable of reasoning, planning, and executing complex tasks autonomously. RAG (Retrieval Augmented Generation) systems to connect your internal data to powerful language models. Voice AI and virtual receptionists available 24/7, capable of handling incoming and outgoing calls. Automation of business processes with n8n and Make, reducing repetitive manual tasks. CRM integrations and intelligent workflows connecting your existing tools. Omnichannel chatbots and AI assistants deployed on WhatsApp, web, and phone.

Notable experiences:

BCG — Development of an automation system for one of the world's largest consulting firms. Holiday Pirates — AI integration to optimize data flows and internal processes. Datacom — Design of AI agent architectures for complex workflows. I also built and deployed an omnichannel AI receptionist in production combining Voice AI and Chat with shared memory, available for live demonstration.

My approach:

I don't deliver tools. I deliver systems that actually work in production and create a measurable impact on your business. Every project begins with a deep understanding of your processes before touching a single line of code or workflow.

I work primarily remotely and am available for short or long-term assignments, in French and English.

If you are looking for an AI engineer who understands both the technical and business challenges, let's discuss your project.

Languages

English
Native or bilingual
French
Fluent

Workplace preferences

Remote only

Primarily works remotely

BCG (Boston Consulting Group)
AI Solutions Consultant
January 2025 - March 2025 (2 months)
▸ Audited consultants' financial document review process, mapping 8–12 hour per-report bottlenecks and identifying root causes in unstructured PDF handling
▸ Architected and built production RAG system parsing 100+ page 10-K/10-Q filings using semantic retrieval, confidence scoring, and multi-stage validation
▸ Engineered citation tracking and hallucination detection layer achieving 95%+ factual accuracy — zero hallucinations on live financial data
▸ Delivered full technical documentation and handover playbook enabling 20+ strategy consultants to operate the system independently
▸ 60% reduction in insight extraction time vs. baseline manual process; system adopted as standard workflow by the consulting team
Multi-Agent Systems Artificial Intelligence Automation RAG API
Andzen
Full-Stack AI Engineer
October 2025 - January 2026 (3 months)
▸ Conducted end-to-end audit of Klaviyo marketing audit workflow, quantifying 8–72 hours of repetitive manual analysis per client engagement
▸ Sole engineer on AuditFlow — a multi-tenant SaaS platform integrating 15+ APIs (Klaviyo, Shopify, analytics) with React dashboard, RBAC, and automated compliance checks
▸ Implemented segmentation analysis engine and AI-generated audit interpretation layer, eliminating manual report writing entirely
▸ Authored client-facing playbooks and operator documentation enabling non-technical staff to manage and interpret audit outputs
▸ Audit time reduced from up to 72 hours to 30 minutes; unlocked 10x throughput; 100+ audits processed monthly in production
Datacom
Data & Automation Engineer
April 2024 - July 2024 (3 months)
▸ Mapped manual data flows across timesheet, payroll, and invoicing systems, quantifying 60+ hours of monthly reconciliation work across teams
▸ Designed and implemented automated ETL pipeline with multi-layer validation catching discrepancies before payroll processing — eliminating downstream errors
▸ Built real-time operational dashboard with automated error alerting, giving finance teams instant visibility into data quality
▸ 80% reduction in manual processing time; 95% fewer billing errors; 60+ hours reclaimed monthly — ROI realized within first billing cycle