AI / ML & Multi-Cloud Solutions for a Better World Architect
Hands-on AI Engineering · Data & MLOps · Multi-Cloud DevOps · SRE · FinOps
Customer-facing architect and hands-on engineer combining AI engineering, enterprise data integration and multi-cloud delivery—from discovery and architecture through prototyping, testing and delivery.
Tell me what you do. What are your pain points? What do you want to improve? What better business outcomes are you aiming for — and why not, what world do you want to build?
Talk to me directly, and we will do it together.
I’m a customer-facing AI solutions architect and hands-on engineer who enjoys turning real business challenges into practical solutions. I work alongside customers and engineering teams from the first conversation through discovery, architecture, building, testing and delivery.
I bring together AI and agentic workflows, enterprise data integration and multi-cloud engineering, using technologies such as RAG, AWS, Azure, GCP, OpenStack, DevOps and MLOps. What matters most to me is finding the right solution for the people who will use it.
Some of my work includes:
In all my projects, I listen to people, understand their pain points and help them build, together, a realistic solution that makes their work easier, faster and more efficient, with better results. This is my passion and my mission. It is what I love to do: not just building products, but building real solutions for human endeavours.
Currently I am working as a Multi-Cloud AI Solutions Architect & Hands-on Engineer at Rackspace Technology.
One person. Three technical strengths. One human-centered mission.
Building AI applications that connect models, knowledge and business workflows — with people in control.
Moving, validating and connecting enterprise data so AI and analytics can use it.
Deploying and operating reliable, cost-aware systems on public, private and hybrid cloud.
The skills that turn technology into outcomes people can trust and use.
How I connect technology to business outcomes — practised in customer workshops and presales at Rackspace, consulting at FPT Software and FinOps at Gameloft.
How I work with people — grounded in years of supporting customers, training engineers and working across countries and languages.
View Résumé PDF Download Résumé PDF
12/2024 – Present
Rackspace Technology, Vietnam
I lead cloud and AI readiness workshops to understand customer pain points, assess team skills and resources, and turn findings into solution architectures and working POCs — combining presales with hands-on engineering across OpenStack, VMware, AWS, Azure and Google Cloud.
07/2024 – 12/2024
FPT Software, Vietnam
I designed enterprise cloud and AI solutions connecting fragmented documents and business systems with the decisions users needed to make.
04/2024 – 05/2024
Accor Group
I improved how developers navigate and understand an enterprise API portfolio, combining documentation architecture with integration analysis.
11/2019 – 04/2024
Gameloft
Within the team operating worldwide mobile-game and digital-marketing infrastructure across the Americas, Europe and Asia — a Montreal private data center plus AWS, Alibaba Cloud, Azure and GCP. My focus included operations, data tooling and FinOps.
2017 – 2019
Soramitsu
12/2015 – 06/2016
Ademax
07/2006 – 12/2015
Polycom
09/2000 – 08/2004
Air Liquide
09/1996 – 09/2000
British Telecom, Paris
AFPA ESIEE, Paris
Evidence of how I work: what each project was for, what I built or designed, and its real delivery stage. Hover or tap a card for a summary.
Private credit and asset appraisal POC with human decision authority
Rackspace Technology · ACB Bank
A private credit and asset appraisal proof of concept that uses specialist agents, document RAG and cited evidence to streamline reviews — while human decision-makers stay in control.
SAP-oriented planning and material readiness for a cosmetics supply chain
FPT Software (architecture) · SPARC portfolio implementation
A copilot design spanning three warehouses across countries to anticipate material shortages, coordinate inventory and support production readiness — with SAP integration and human approval.
Developer tooling and DevSecOps POC — AI-assisted SDLC with human oversight
Personal project
A 25-stage prototype of an AI-assisted software lifecycle — requirements, architecture, coding, test automation, security reviews, CI/CD and incident root-cause analysis — with people reviewing each step.
Community lab: full-stack agent and ontology platform
Rackspace Technology — internal
An internal customer-zero platform for developing and demonstrating reusable AI capabilities for Rackspace teams.
EV fleet and mobility intelligence demo lab
Personal project
A demo lab for electric-vehicle fleets: fleet command and vehicle health, charging network and route intelligence, ride-demand forecasting, and an ADAS model-improvement loop — “Drive, Observe, Learn, Predict, Optimize, Improve”.
Model-aware GPU VM sizing for LLM inference
Rackspace Technology
A model-to-GPU VM sizing catalog that helps select GPU configurations from model size, precision, serving framework, context length and concurrency — to support configuration and cost discussions.
Ready-to-use agent marketplace for business verticals
Rackspace Technology
A no-code, ready-to-use agent library for banking, IT and e-commerce that helps customers discover AI use cases and scope POC, MVP and production pathways — supporting onboarding to Rackspace’s GPU-for-rent offering.
Rural AI platform POC — two copilots on one shared platform
Personal project · independent demo
A working POC with two AI copilots: a Rural Finance Copilot for agricultural micro-credit, and a Project Monitoring & Impact Copilot for rural development projects — with cited answers and human approval.
Human-governed, Web3-verifiable autonomous AI agents
Personal project
An AI treasury agent investigates supplier invoices inside a system where it has no authority to move money: policy decides, humans approve, and a blockchain makes the audit record tamper-evident.
AI troubleshooting built on Kepner-Tregoe problem analysis
Personal project
A support-ticket workflow built on a 10-step troubleshooting action plan rooted in Kepner-Tregoe problem analysis, with an AI agent that finds the shortest path to resolution from the knowledge base, customer history and problem register.
MCP runtime security and guardrails POC
Personal project
A prototype guardrail layer for AI agents: policy evaluation, risk checks, human escalation, audit, and detection of unauthorized tool use.
Understand meaning before predicting language
Personal research project
A research POC that asks what is happening, who is involved and what information is missing — then scores a Transformer's next-token candidates against that structure.
See what an action changes before taking it
Personal research project · builds on MLM
Built on the Meaning Language Model, ACM predicts what an action does: its consequences for the actor, for those it affects and for other actors’ goals — across a blast radius from the place it happens out to society and the earth.
OSPC-to-Flex OpenStack migration workflows
Rackspace Technology
Migration workflows for legacy OpenStack, VMware and hyperscaler workloads moving to Rackspace Flex OpenStack — covering VM migration, validation, rollback and cross-region DR.
~US$40K/month cloud savings and faster ETL for a global gaming estate
Gameloft — contribution within the infrastructure team
Production cloud, data engineering, reliability and FinOps work for worldwide mobile-game and digital-marketing infrastructure across a private data center, AWS, Alibaba Cloud, Azure and GCP.
Systems, workflows and reference architectures I designed or contributed to across AI, data, banking, multi-cloud, MLOps and supply chain. Each one is labelled with its real status — most are reference or proposed designs, not deployed systems.
Full-stack target design for an enterprise banking copilot: SSO and RBAC, Flowise Agentflow V2 agents behind an MCP tool gateway, durable workflows, private LLM inference, GitLab MLOps and AWS staging.
Credit-appraisal AI runs on a private OpenStack Kubernetes cluster; GitLab is the MLOps control plane; AWS SageMaker handles training, evaluation and a staging replica. Only approved, de-identified data moves to AWS.
The same hybrid pattern on Google Cloud: production inference stays on OpenStack while Vertex AI Pipelines, a model registry and a GKE staging replica support development and evaluation.
A ten-step improvement loop that keeps live customer data on private OpenStack and moves only approved, de-identified datasets to AWS for training and staging.
The ten-step improvement loop on Google Cloud: approved, anonymized data goes to Cloud Storage, Vertex AI trains and evaluates, GKE stages, and releases return to OpenStack.
End-to-end MLOps on AWS: governed data, SageMaker training and evaluation, GitLab CI/CD with security checks, staged rollout and monitored retraining — with no automatic promotion to production.
A secure, highly available internet-banking platform on AWS: edge protection, multi-AZ application and data tiers, hybrid connectivity to core banking, and built-in operations and security.
Concept for AI-assisted rural micro-credit: farmers upload photos and documents, specialist agents analyse them, and a loan officer makes the final decision before core-banking integration.
Private, human-centred supply-chain AI for one factory, three warehouses and SAP S/4HANA: an agent hierarchy recommends, people approve, and SAP execution stays authorised and traceable.
Professional certifications are labelled “Certification”; courses and programmes are labelled “Course”.
AWS Certified Solutions Architect – Professional
AWS Certified Cloud Practitioner
Google Cloud Professional Cloud Architect
HashiCorp Terraform Associate
Cisco networking — CCNA, CCNP, Firewall
Dataiku Machine Learning
Deep Learning Specialization
Machine Learning in Production
MLOps Platforms: SageMaker & Azure ML
Generative AI
Software Architecture Case Studies
Advanced Python
REST API Engineering
Blockchain Fundamentals
Kepner-Tregoe Certified Problem Solver
Learning Science Techniques
MBA Essentials
Technology Leadership in the Age of AI
Telecom engineering
Open to conversations about AI, data and cloud solution architecture — remote or hybrid. Email is the fastest way to reach me.
Ho Chi Minh City, Vietnam
