Enias Cailliau

Teseo

Co-founder & CTO

Steamship

Put language AI into the hands of every developer.

Founding Engineer

Litto

* Productionise AI solutions for a 24/7 SaaS platform
* Develop MLOps pipelines including data engineering, data quality, and automated training
* Motivate and cultivate a team of engineers

Radix

* Lead the design of machine learning architectures  
* Supervised a successful master dissertation 
* Coached interns to become excellent machine learning engineers

Technical Lead in Machine Learning

* Develop and deploy end-to-end Machine Learning models to solve customer problems.
* Natural Language Processing - Knowledge extraction, Semantic search, Recommender engines.
* Big Data Engineering - Data Pipelines, ETL, Data Governance
* Machine Learning Operations - Building Pipelines and Maintenance.

Machine Learning Engineer

I help developers & entrepreneurs build with GPT | Founding Engineer at Steamship

Turned down Facebook to build winning startups. 

- Build what customers want.

The power of compounding.

elitizon

Co-Founder and CTO of a technology venture studio.

- Crafting AI strategies aligned to business goals
- AI/ML Strategy and Implementation
- AI Engineering
- Ensuring models deliver repeatable value in production
- Advanced Data Engineering

CTO - ELITIZON Ltd

Quantmetry - Part of Capgemini Invent

Data and AI Architect consultant for "The State-of-the-Art AI company"

Data & AI - Architect Consultant

Decathlon 

- Data Strategy roadmap
- Data Governance design
- Data Engineering (AirFlow / Spark / AWS EMR / Databricks / Scala / Python )
- Data and Analytics Operating Model design
- Modern Data Stack Roadmap
- Cloud data architecture (AWS / GCP)
- AI/ML architecture and DataOps
- Data Factory Ingest - Tech Lead

La French Tech Hong Kong - Shenzhen

Helping French-Startups Ecosystem in HK / China, Great Bay Area

Board Member

WaveX

🌊 Using wave energy to produce electricity while reducing carbon emissions and combating erosion on coastlines

Technical Advisor

Air Liquide

2018 - 2020 : Architect, Design, Develop & Operate a Worldwide Analytics Solution on Microsoft Azure to optimise the Industrial Gases Supply Chain  | Azure DataFactory ETL | Azure DWH Synapse | Azure Cubes | DataViz PowerBI 

2008 - 2020:  Architect, Design, Develop and Maintain a mission-critical global solution to manage and optimise industrial Gas deliveries by truck on 3 continents ( America  | Europe  | Asia  ) - South Africa deployment planed in 2020  | Transport Management System

✅ 2 Billions Euros Business / year
✅ 1.5 millions Deliveries by Truck / year
✅ 40 000 Customers served 

🛰 Real Time Visibility of the Supply Chain
📊 Real Time Inventory at the customer site (Internet Of Things)
📈 Consumption Forecasting
🤖 Vendor Management Inventory | Operation Research & Tactical & Operational Optimisation 


2010 - 2012:  Architect, Design, Develop - Internet of Thing Solution for patients monitoring (Sentinel)

✅ > 100K Patients monitored in France

2016 - present: Signed a strategic Digital Transformation Partnership with L’Oréal Group (Fortune 500) to create and develop embedded softwares and artificial intelligence cloud infrastructures for connected beauty devices 

✅ 2 Connected Beauty Solutions presented at CES Las Vegas (2017 & 2020)
✅ New Beauty Digital Business - Color & Co - L'Oréal USA

2013: first AWS powered Cloud Solution deployed within Air Liquide Group. Adoption of a Cloud First strategy for application development

Founder & CEO at CALIATYS | Air Liquide Group

Winwise

⛰   2000 to 2008 : Developed of one of the leading training and expertise center for software development on Microsoft Technology in Paris (Microsoft .NET |  MS Bitzalk )

👏  2001 : Development of one of the first e-Commerce Solution (connected to SAP) - Spare parts for Sidel Tetra Pack

👏   2002 to 2008 : consulting and expertise to build a new generation heath & life insurance information system (Le Gan | Groupama - Lausanne) 

💵  2000 : Received 1.5 M Euros of founding 

👏  2000 : Signed a consulting contract for AXA Bank - Development of a credit/risk management system

🌱  2000 : Signed a partnership with Microsoft France to train Microsoft Consulting Team and Microsoft Partners on Microsoft Biztalk, Web Services and XML technologies

👉 1998 : Signed a partnership with VALTECH to develop a software expertise practice on Microsoft
 
🚀  1998 : Launched an expertise and consulting spinoff of Winwise Group valued 3.8 M Euros after 2 years. 

Winwise Group was the leading pure player software development company specialized in Microsoft Technologies in France from 1995 to 2008. Winwise was acquired by "ALTEN Group" in 2008.

Business Offer

➡ Microsoft Technology Consulting
➡ Custom Software Development
➡ Innovation Consulting
➡ Software architecture consulting / Audit  / Technical Expertise

🎖 Microsoft Gold Partner
🎖 Microsoft Certified Training Center / Prometric Certification Center


✅ 5K Software engineers  trained

✅ Software expertise missions for:

  ☑ Microsoft France
  ☑ ATOS
  ☑  Business Object
  ☑ Schneider Electric
  ☑ Alcatel
  ☑ Dassault Système
  ☑ Le GAN (Insurance)
  ☑ Sidel / Tetra Pack

Founder & CEO BUSINESS DESIGN CONSULTING | Winwise Group

Microsoft Conference Speaker 

Consulting & Expertise Missions for CORIOLIS Telecom | Schneider | Air Liquide | VALTECH | ALCATEL | D'ARVA | Ministère des affaires étrangères | LECTRA SYSTEM | MICROSOFT

🎖 Microsoft Certified Trainer
🎖 Microsoft Certified Solution Developer
🎖 Microsort Certified Database Administrator (SQL Server)

Software architecture consulting on Microsoft Technology stack.  IT Trainer on Microsoft Technologies

Microsoft Technology Consultant

Banque Populaire

Lead developer of a transaction oriented service bus (ESB)  to implement of a service oriented architecture (SOA) | System & Network programming in C++ 

✅ Distributed architecture deployed in 200 Bank Agencies
✅ Backbone of one of the first Web e-Banking solution in France

Software Developer

On a Mission to Democratize Data Management & AI 🚀

As a seasoned data and AI strategist 🧠, I am on a mission to democratize data management and AI. 

👉 I empower organizations to build robust, scalable data platforms that drive transformative business outcomes. 

👉 With over 20 years of global technology leadership experience, I offer strategic guidance and hands-on expertise to help companies fully leverage the power of data and AI.


My consultative approach begins with deeply understanding your unique requirements and goals. I then craft a tailored roadmap to build, optimize, and scale your data infrastructure and capabilities. Whether you’re looking to transition to a data-centric business model, enhance your AI/ML initiatives, or overcome specific data challenges, I provide the specialized skills and insights you need.

👉 My services include:

-  Architecting scalable, resilient data platforms and pipelines
-  Streamlining deployment and management in the cloud
-  Implementing cutting-edge machine learning models
-  Coaching teams on data engineering and AI best practices
-  Devising data strategies aligned to business objectives
-  Overseeing DevOps, CI/CD, and MLOps processes

As a startup founder and advisor, I recruit and lead high-performing engineering teams. My passion is empowering people and organizations to maximize the value of their data. I have contributed to over €2B in revenue through data-driven software products and successfully grown and sold a company for 140x its initial investment.

With deep expertise across data engineering, machine learning, and product development, I am eager to collaborate with clients to create data-centric cultures and power AI innovation. Let's connect to discuss how we can lead the future together.

Currently, I spearhead a startup studio in Hong Kong and offer consulting services to clients around the globe. My expertise has been sought by Fortune 500 companies, and I am passionate about helping businesses unlock the full potential of their data, amplifying its value and impact.

👉 When you're ready to elevate your data strategy, I'm just a click away: 📆 

https://topmate.io/raphael_mansuy 

In the ever-evolving landscape of data engineering, data science, and AI, staying ahead of the curve is crucial. I share daily insights, tips, and tricks to help you stay informed and inspired.

Let's connect and lead the future together. Follow me for the latest updates and insights. 😎. ↗️ 🔔

Raphaël MANSUY

Scaling Graph Chain-of-Thought Reasoning: A Multi-Agent Framework with Efficient LLM Serving

Making Graph Reasoning Work: Multi-Agent Architecture Cuts Costs by 95% While Improving Accuracy ...

Can an AI answer "What's the average price of items often viewed with product A?" Most current systems either hallucinate the answer or burn through thousands of tokens trying to figure it out. Here's why—and what changes when we fix it.

WHY This Matters

When LLMs need to reason over structured data like knowledge graphs, they face a dilemma. Traditional retrieval methods treat graphs like flat text, losing all the connections between entities. Chain-of-thought reasoning helps, but it creates a new problem: each reasoning step re-processes the entire context, leading to:

👉 Token costs exceeding $3 per query on commercial models
👉 Accuracy dropping below 50% on multi-hop questions
👉 Latency stretching to 40 seconds per answer

The core issue? Single-agent architectures that stuff everything into one massive prompt, causing the "lost-in-the-middle" problem where crucial information gets buried.

WHAT GLM Does Differently

GLM introduces a multi-agent framework where specialized agents handle distinct tasks:

👉 A classifier determines if a question needs simple lookup or complex reasoning
👉 A reasoning agent maintains a lightweight "notebook" of relevant facts
👉 An action agent generates executable Python code instead of verbose reasoning chains
👉 A graph retriever executes these snippets and returns precise results

Instead of maintaining one enormous context, each agent operates on minimal, task-specific information. The system replaces 8+ LLM calls with 2-3 focused interactions.

HOW It Achieves Better Performance

The technical innovations work in concert:

Vertex-centric caching groups node information with its immediate neighbors, creating reusable chunks that persist across queries. When multiple questions reference the same entities, the system reuses computed representations instead of recalculating them.

Priority-based eviction keeps frequently-accessed graph data in memory while discarding temporary reasoning artifacts. This prevents the cache from filling up with information that won't be reused.

Pipelined execution overlaps graph retrieval with LLM inference. While the model generates code, the system preemptively fetches data, hiding latency that would otherwise stack sequentially.

The results speak clearly: 38% higher accuracy, 95% fewer tokens, 90% lower latency, and 15× higher throughput compared to existing graph reasoning systems.

This matters because structured data powers most enterprise applications—from recommendation engines to medical diagnosis systems. Making graph reasoning both accurate and affordable opens these capabilities to practical deployment.

Data Engineering | DataScience | AI & Innovation | Author | Follow me for deep dives on AI & data-engineering

Scaling Graph Chain-of-Thought Reasoning

AI Product Manager @ Google | Open Source Awesome LLM Apps Repo (#1 GitHub with 78k+ stars) | 3x AI Author | Views are my Own

We just ran the biggest AI Agent course ever 🤯

Here's the list of all the resources we released from Day-1 to Day-5.

250+ pages of production AI Agent insights. 100% free.

📌 Day 1: Introduction to AI Agents (50 pages)
Most agent demos fail in production. This covers the 5 levels of AI agents, from basic reasoning to self-evolving systems, plus the security and architecture patterns that actually work at scale.

Link: https://lnkd.in/dQSzVjR7

📌 Day 2: Agent Tools and MCP (50 pages)
MCP servers can add new tools without your approval. Your poetry bot suddenly gains purchasing power overnight. This breaks down tool design, the Model Context Protocol, and the security risks hiding in plain sight.

Link: https://lnkd.in/dUztxWDA

📌 Day 3: Context Engineering and Memory (70 pages)
Agents with zero long-term memory treat every session like meeting you for the first time. Learn how to build memory systems that actively curate context, not just save conversations.

Link: https://lnkd.in/dMqjwrqF

📌 Day 4: Agent Quality (50 pages)
Traditional QA doesn't work for agents. You can't unit test judgment. This covers the 4 pillars of agent quality (effectiveness, efficiency, robustness, safety) and how to actually evaluate the reasoning process.

Link: https://lnkd.in/dQR3_HkG

📌 Day 5: Prompt to Production (40 pages)
Building an agent takes 2 weeks. Trusting it in production takes 6 months. Learn evaluation gates, circuit breakers, and evolution loops that close the gap between demo and deployment.

Link: https://lnkd.in/dVHSeWmE

With insights from experts including Antonio Gulli, Alan Blount, Michael Zimmerman, Vladimir Vuskovic, Michael S., Kanchana Patlolla, Santiago (Sal) Diaz, Madhurranjan Mohaan, Kimberly Milam, Meltem Subasioglu, Turan Bulmus, Wafae Bakkali, PhD, Dr. Sokratis Kartakis, Gabriela Hernández Larios, Ran Li, Elia Secchi, and Huang Xia 🔥

Kudos to Kanchana Patlolla and Anant Nawalgaria for the amazing organization 🙌

The best part?
250+ pages. Zero fluff. 100% free.

Repost this so your network doesn't miss it ♻️

taplio

The Battle

Enias Cailliau

Raphaël MANSUY

Followers

Posts

Engagements

Likes

Performance ScoreRaphaël MANSUY's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Enias Cailliau

Raphaël MANSUY

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Raphaël MANSUY's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreRaphaël MANSUY's LinkedIn Performance