Modernization Strategy
Why modernization is not simply refactoring, and how to choose the right path.
Sylvain Leroy
Cloud & Application Modernization Architect
I help software companies, banks and enterprise teams migrate, modernize and optimize complex platforms across AWS, GCP and cloud-native environments.
20+ years in software modernization · AWS & GCP certified · SaaS, banking, insurance and enterprise platforms
Cloud and Application Modernization
Field notes on cloud & modernization
I write about cloud migration, modernization strategy, legacy transformation, architecture complexity, performance optimization and the real-world constraints of enterprise IT.
Architecture & Resilience
How complexity impacts RTO, operations and business continuity.
Why RTO explodes when architecture complexity grows
Read the articleCloud Engineering
Practical lessons from AWS, GCP, data platforms and cloud-native systems.
Performance optimization for AWS Glue using Microsoft SQL Server Spark connector
Read the articleAI Agents
Building an AI-powered diagnostic agent for real-world industrial equipment.
How to build an AI-powered diagnostic agent for industrial equipment
Read the articleLegacy Modernization
Strategies for moving PowerBuilder applications into the cloud era.
Modernizing PowerBuilder Applications: Strategies for the Cloud Era
Read the articleHybrid Cloud
Monitoring hybrid VPN connectivity between Azure and AWS in production.
Monitoring Hybrid VPN Connectivity Between Azure and AWS
Read the articleFrom the blog
Latest articles
Fresh field notes on cloud, modernization, data engineering and AI — published as I go.
artificial-intelligence
AionUi: A GUI That Wraps Your CLI AI Agents Instead of Replacing Them
AionUi wraps CLI AI agents like OpenCode and Claude Code — an honest review of the open-source GUI that respects the agent ecosystem instead of reinventing it.
artificial-intelligence
Basic Memory: Giving My AI Assistant a Memory That Lives in My Obsidian Vault
basic-memory gives my AI assistant persistent memory as plain Markdown synced to my Obsidian vault — transparent, editable, and a real accuracy boost. Part 6 of the local-AI stack series.
artificial-intelligence
Building a Private Local-AI Stack: Why and How
How I built a private local-AI stack that automates real work — email triage, recall, drafting — while keeping sensitive data on my own hardware. Part 1: the use case and the sensitivity boundary that drives every choice.
artificial-intelligence
LightRAG as a Personal Knowledge Base for AI Agents
How I run LightRAG as a local semantic index over my personal documents, wired into AI agents via MCP — the architecture, the tuning, and the silent outage that taught me health checks lie. Part 2 of the local-AI stack series.
artificial-intelligence
OpenRouter and the Case for Safe Models
One API, many models, and provider pinning: why OpenRouter sits at the non-PII half of my local-AI stack, and how I draw the line that sensitive data never crosses. Part 3 of the local-AI stack series.
artificial-intelligence
Right-Sizing a Local LLM to Your Laptop with llmfit
The PII half of my stack runs a local model — but which one fits 16 cores and 20 GiB? llmfit scores every model against your actual hardware so you stop guessing. Part 4 of the local-AI stack series.
Gallery
Moments from the field
Ready to start?
Need help modernizing a platform, preparing a cloud migration, or reviewing your architecture?
Let's talk about your platform, your constraints, and where AWS, GCP or cloud-native architecture can move the needle.
Get in Touch