Stop Burning Tokens: A Practical Guide to Using Claude and Claude Code Efficiently
How business users and software developers can get better results while spending fewer tokens
Services
How I help teams build AI products
AI Products & Agents
Production AI agents, copilots, RAG systems, and workflow automation built for reliability, security, and real-world use.
Learn more →
Cloud Platforms for AI
AWS platforms that power production AI—from infrastructure and CI/CD to observability, security, and scalable cloud architecture.
Learn more →
AI-Native Engineering
Modern software delivery with coding agents, specification-driven development, developer tooling, and engineering workflows that scale.
Learn more →
Technical Leadership
Hands-on engineering leadership for ambitious teams—designing architecture, solving difficult technical problems, and shipping production software.
Learn more →
Open Source
Charmant.AI: a safer way for AI agents to use local tools.
Charmant.AI is an open-source local AI agent I built for working with files, shell commands, logs, and repeatable system workflows.
The project explores a practical production problem: how to give an AI agent useful capabilities without giving it unrestricted access to the machine. It combines explicit permission controls, command risk classification, reusable skills, project instructions, and structured audit logs.
What it includes
- Local agent tools for files, shell commands, logs, and system inspection
- Configurable permission modes for human-approved automation
- Dynamic shell-command risk classification before execution
- Reusable skills, project instructions, and structured audit logs
Writing
Engineering insights
Migrating Amazon ECR Images Between AWS Accounts with a Simple Bash Script
A lightweight Bash script for migrating Amazon ECR images between AWS accounts. The script uses Docker Buildx to copy image references directly between registries, preserving tags and multi-architecture manifests without requiring a local pull, tag, and push workflow.
Building AI-Native Engineering Teams Without Losing Engineering Discipline
AI-native engineering is not about replacing engineers with prompts. It is about redesigning the development system so small, senior teams can use AI across planning, design, implementation, testing, review, and operations without losing control of architecture, quality, or security. This article explains how founders and technology leaders can combine Domain-Driven Design, spec-driven development, TDD, curated context, and automated guardrails to ship products in weeks rather than months while keeping the codebase maintainable and production-ready.
No Backend Needed: Running Python in React with Pyodide
A step-by-step guide to integrating Pyodide into a React + TypeScript + Vite app, loading Python packages like NumPy and Matplotlib in the browser, and generating client-side data visualizations from CSV revenue data.
About
Who I build for
I work with startups and engineering teams building AI products, developer tools, and cloud platforms. Whether I join an existing team or lead a critical project, my focus is the same: turning ambitious ideas into reliable production software.
- Teams building AI products for customers
- Engineering teams adopting AI-native development
- Companies moving AI systems from experimentation into production
- Founders who need senior engineering experience without building a full architecture team
Track record
Selected work
- Built production AI systems with security, auditability, and operational controls designed in from the start.
- Designed AWS platforms that balance scalability, resilience, observability, security, and cost.
- Helped engineering teams adopt AI-native development using coding agents, specifications, testing, and curated project context.
- Built production software for regulated industries, from secure cloud platforms to developer tools and AI systems.
Speaking & publications
Talks & workshops
Practical sessions on AI engineering, cloud architecture, security, and building production-ready systems.
- Reconstructing the Past with Knowledge Graphs: Empowering Historians through Neo4j and GenAI (FOSSASIA Summit 2026)
- Unlocking the Power of Graph Databases with Neo4j (FOSSASIA Summit 2025)
- Connected: Solving Real-Life Problems Using Python and Graph Databases (PyCon ID 2023)
Research & writing
In-depth work on AI systems, agent security, cloud platforms, and AI-native engineering practices.
- Building AI-Native Engineering Teams Without Losing Engineering Discipline →
- Zero Trust Security Architecture for AI Agents in Cloud Environments (Cyber Security: A Peer-Reviewed Journal, Coming soon)
- Differences between traditional network security and security in the cloud (Cyber Security: A Peer-Reviewed Journal, 2025) →
- Node.js Security (Cyber Security: A Peer-Reviewed Journal, 2017) →
Ready to build?
Whether you need a second opinion on an architecture, help shipping an AI system, or an experienced engineer to strengthen your team, let’s start a conversation.