# Jean-Baptiste Terrazzoni

> Problem solver, mostly with code. I build small apps, put AI in the loop, and write about what I learn.

Jean-Baptiste Terrazzoni (problem solver) builds small apps and tools, and writes notes on software, AI agents, and owning your own stack. Articles are available in English and French (French under /fr/).

## Articles

- [Your next AI skill is worldbuilding](https://www.jterrazz.com/articles/24-your-next-ai-skill-is-worldbuilding): Stop asking which model. Start asking what world. A practical guide to building working worlds around AI: identity, tools, skills, knowledge, context, and feedback.
- [SigNews: Mapping global news narratives with AI](https://www.jterrazz.com/articles/13-signews-mapping-global-news-narratives-with-ai): SigNews, a dashboard concept for monitoring global news narratives. AI distills signal from noise, mapping where stories align, clash, and fade into speculation.
- [When AI becomes the product: Designing AI-native systems](https://www.jterrazz.com/articles/23-when-ai-becomes-the-product-designing-ai-native-systems): Design intelligent systems that blend traditional code with AI reasoning. Learn architectural patterns for building applications where intelligence is a first-class component.
- [Autonomous AI agents: When to delegate entire workflows](https://www.jterrazz.com/articles/22-autonomous-ai-agents-when-to-delegate-entire-workflows): Explore autonomous AI agents that work independently on complex tasks. Learn when to delegate entire workflows and how to design systems for agent autonomy.
- [Directing AI agents: From writing code to orchestrating it](https://www.jterrazz.com/articles/21-directing-ai-agents-from-writing-code-to-orchestrating-it): Learn to direct AI agents as a developer. Transform your workflow from writing code to orchestrating intelligent agents that handle implementation details.
- [The four levels of AI mastery](https://www.jterrazz.com/articles/20-the-four-levels-of-ai-mastery): A framework for understanding AI integration levels in any field. From simple assistants to fully programmable intelligence, map where you are and where you're heading.
- [Cursor IDE in practice: Agent mode and intent-driven coding](https://www.jterrazz.com/articles/19-cursor-ide-in-practice-agent-mode-and-intent-driven-coding): Master Cursor IDE for AI-assisted development. Learn context-aware coding, agent mode, and intent-driven development to compress mechanical work into high-level direction.
- [What remains when machines handle everything?](https://www.jterrazz.com/articles/17-what-remains-when-machines-handle-everything): When work stops being about survival, it becomes expression. Exploring how AI abundance transforms careers, identity, and what it means to contribute in a post-execution world.
- [The collapse of the middle: Where value goes when AI executes](https://www.jterrazz.com/articles/16-the-collapse-of-the-middle-where-value-goes-when-ai-executes): The comfortable middle layer of knowledge work is evaporating. As execution becomes cheap, value concentrates at the extremes: pure judgment at the top, pure humanity at the bottom.
- [How to actually use AI](https://www.jterrazz.com/articles/15-how-to-actually-use-ai): A practical guide to working with AI. Stop treating it as a last resort and start directing it like a team of specialists: researcher, writer, analyst, and strategist on demand.
- [Your moat is melting: What AI does to knowledge work](https://www.jterrazz.com/articles/14-your-moat-is-melting-what-ai-does-to-knowledge-work): The execution layer that pays your salary is collapsing. As AI commoditizes knowledge work, discover what actually makes you valuable in an age of abundant intelligence.
- [Clean architecture in practice: Entities, use cases, and layers](https://www.jterrazz.com/articles/12-clean-architecture-in-practice-entities-use-cases-and-layers): Deep dive into Clean Architecture. Organize your code into entities, use cases, adapters, and drivers to create systems where business rules remain untouched by technical changes.
- [Hexagonal architecture: Separating business from technology](https://www.jterrazz.com/articles/11-hexagonal-architecture-separating-business-from-technology): Explore hexagonal architecture and ports & adapters pattern. Learn to isolate your business logic from external concerns like databases, APIs, and frameworks.
- [The dependency rule: Structuring code for independence](https://www.jterrazz.com/articles/10-the-dependency-rule-structuring-code-for-independence): Master dependency management in software architecture. Learn how to structure code so dependencies flow in the right direction, keeping your core logic independent.
- [Maintainable software: Building applications that last](https://www.jterrazz.com/articles/9-maintainable-software-building-applications-that-last): An introduction to building maintainable software. Learn the foundational principles that help applications evolve gracefully over time without accumulating technical debt.
- [What if we got online shopping wrong?](https://www.jterrazz.com/articles/8-what-if-we-got-online-shopping-wrong): A vision for reimagining e-commerce as an open, decentralized protocol. Exploring how to bring back community, transparency, and fair competition to online shopping.
- [Building a web server with TypeScript and Koa](https://www.jterrazz.com/articles/7-building-a-web-server-with-typescript-and-koa): Build a production-ready web server with TypeScript and Koa. Learn to set up routing, middleware chains, and type-safe request handling in a lightweight framework.
- [Building an expert system in Python (backward chaining)](https://www.jterrazz.com/articles/6-building-an-expert-system-in-python-backward-chaining): Build an expert system from scratch in Python. Implement a backward-chaining inference engine that reasons with rules and facts to solve logical problems.
- [X86-64 assembly basics: Writing your first functions](https://www.jterrazz.com/articles/5-x86-64-assembly-basics-writing-your-first-functions): Get started with x86-64 assembly language. Set up your toolchain, understand registers and calling conventions, then write and link your first assembly functions.
- [Building a program that writes itself (a quine)](https://www.jterrazz.com/articles/4-building-a-program-that-writes-itself-a-quine): Explore the fascinating world of quines, programs that output their own source code. Learn the theory behind self-replicating code and implement quines in multiple languages.
- [Building nm and otool: Parsing Mach-O binaries in C](https://www.jterrazz.com/articles/3-building-nm-and-otool-parsing-mach-o-binaries-in-c): Learn Mach-O binary format by building nm and otool in C. Decode symbol tables, parse headers and sections, and understand how macOS executables work.
- [SHA-256 and MD5 in C: Implementing hash functions from scratch](https://www.jterrazz.com/articles/2-sha-256-and-md5-in-c-implementing-hash-functions-from-scratch): Implement SHA-256 and MD5 hash algorithms from scratch in C. Master bitwise operations, message padding, compression loops, and validate with test vectors.
- [Writing your own malloc in C: A memory allocator from scratch](https://www.jterrazz.com/articles/1-writing-your-own-malloc-in-c-a-memory-allocator-from-scratch): Build your own malloc in C from scratch. Learn mmap, memory buckets, first-fit allocation, block coalescing, and alignment to truly understand dynamic memory.

## Experiments

- [Spwn](https://www.jterrazz.com/experiments/spwn): The operating system for autonomous agent worlds. Compose, spawn, ship.
- [PhotoPack](https://www.jterrazz.com/experiments/photopack): Pack your photo library tight: deduplicate, organize, compress.
- [Capitaine](https://www.jterrazz.com/experiments/capitaine): A day to day copilot for your life.
- [SigNews](https://www.jterrazz.com/experiments/signews): Your personal intelligence bureau, scanning global signals 24/7.
- [Jterrazz](https://www.jterrazz.com/experiments/jterrazz): Personal space for experiments, articles, and projects.
- [Typescript](https://www.jterrazz.com/experiments/typescript-packages): A collection of typescript packages for building my node.js applications.
- [Defy Dy](https://www.jterrazz.com/experiments/defy-dy): Synthetic asset platform leveraging Chainlink oracles.
- [CVT Crowdfunding](https://www.jterrazz.com/experiments/cvt-crowdfunding): Blockchain-based crowdfunding platform using CVT tokens.
- [Infrastructure](https://www.jterrazz.com/experiments/infrastructure): Infrastructure as Code for my personal cloud environment.
- [Forge](https://www.jterrazz.com/experiments/forge): Command center for autonomous AI agents and automations.
- [Hoverfly LSP](https://www.jterrazz.com/experiments/hoverfly-lsp): A language server that brings autocomplete, validation, and navigation to Hoverfly API simulation files.
- [Design System](https://www.jterrazz.com/experiments/design-system): A cross-product design system: shared tokens and platform-native components used across my apps.
- [VPN](https://www.jterrazz.com/experiments/vpn): A self-hosted WireGuard VPN running on a TP-Link pocket router flashed with OpenWrt.
- [Expert System](https://www.jterrazz.com/experiments/expert-system): Backward chaining rule based system in Python. RPN, Tree resolver, Tree representation, logic rule system, prompt.
- [Override](https://www.jterrazz.com/experiments/override): Advanced binary security experiment - Exploits and reverse engineering - x86 and x86-64 binaries.
- [Rainfall](https://www.jterrazz.com/experiments/rainfall): Intermediate cybersecurity experiment, and an introduction to binary analysis - X86 binaries.
- [Snowcrash](https://www.jterrazz.com/experiments/snowcrash): Entry level cybersecurity experiment - X86 binaries.
- [Corewar](https://www.jterrazz.com/experiments/corewar): Virtual machine simulating basic processors operations, with basic notions of Assembly. Written in C.
- [NM Otool](https://www.jterrazz.com/experiments/nm-otool): Nm and otool implementation in C. Parse Mach-o files, supports fat binaries, archives and corrupted binaries.
- [MD5 & SHA256](https://www.jterrazz.com/experiments/md5-sha256): OpenSSL implementation in C. Supports md5, sha1, sha256, sha224, sha512 and sha384 algorithms.
- [Malloc](https://www.jterrazz.com/experiments/malloc): C implementation of the malloc library using mmap.
- [Dr Quine](https://www.jterrazz.com/experiments/dr-quine): This experiment addresses auto-replication issues, and confronts you the Kleene's recursion theorem.
- [Fillit](https://www.jterrazz.com/experiments/fillit): Algorithmic C experiment for the 42 Paris School
- [Ft Printf](https://www.jterrazz.com/experiments/ft-printf): Custom printf implementation in C.
- [Get Next Line](https://www.jterrazz.com/experiments/get-next-line): Read lines from a file descriptor.
- [Lem In](https://www.jterrazz.com/experiments/lem-in): Handle a virtual anthill efficiently.
- [Libft ASM](https://www.jterrazz.com/experiments/libft-asm): Basic functions implemented in Assembly using the x86 Intel syntax.
- [Libft](https://www.jterrazz.com/experiments/libft): The basic libc library used in all the 42 school experiments.
- [Push Swap](https://www.jterrazz.com/experiments/push-swap): Sorts data on a stack, with a limited set of instructions and limited number of moves. In C.
- [Piscine Reloaded](https://www.jterrazz.com/experiments/piscine-reloaded): My 42 paris school entry test.

## Optional

- [Photography](https://www.jterrazz.com/photographs): Photography portfolio
- [RSS feed](https://www.jterrazz.com/feed.xml): Article feed
- [GitHub](https://github.com/jterrazz): Open-source work