Going AI-Native: Rethinking How I Develop (Part 1)

 


    I have been working on shifting my own development workflow to what's being called "AI-native" development recently, using Spec-Kit, Spec Driven Development (SDD), and Claude. This is the first post in a series where I'll walk through what I learned from two courses I took on this, what I've actually built with it, and where I'm planning to take it next.

Why I Got Into This

    Since the AI coding tools took the stage, the process of coding became so fast. But the processes around that code, planning, review, testing, deployment, haven't caught up at the same pace. I kept running into this gap myself: Claude could generate a feature in minutes, but I was still planning, reviewing, and testing it the old way. That mismatch is basically what pulled me into learning this properly instead of just using AI tools ad hoc. Also my company started adapting the AI Native development. Then it is only natural that I choose this way.

To get a structured understanding of it, I took two courses:

  • A Udemy course called "Become an AI Native Engineer with Spec-Kit, SDD, and Claude"

  • "The AI-Native SDLC Playbook" from Anthropic Academy

This post covers the core ideas from both, not the course content itself, but what the ideas actually mean and why they matter.

What Is AI-Native Development?

    AI-native development isn't just "using Claude to write code." It's restructuring the entire software development lifecycle (SDLC) around AI doing meaningful work at every stage, not just the build step.

The traditional SDLC has six stages: plan, design, build, test, deploy, maintain. In a traditional setup:

  • Plan: requirements are gathered by committee, distilled through workshops and sign-offs, written up by hand.

  • Design: specs and user stories are written by analysts, then parsed by designers

  • Build: tests and code are handwritten, documentation comes after the fact

  • Test: QA happens at fixed stage boundaries

  • Deploy: humans review every line of code, governance happens in review cycles, often inconsistently

  • Maintain: humans watch production for bugs

In an AI-native setup, each of those stages shifts:

  • Plan: an AI agent synthesizes pain points straight from source material into an intent.md file that's both human readable and machine actionable

  • Design: requirements and design get compressed into one working session with an agent, guided by standards encoded as skills, versioned in Git

  • Build: tests and code are generated by AI, and institutional knowledge is maintained as versioned, machine-readable CLAUDE.md files and skills

  • Test: continuous evals are woven through implementation instead of sitting at fixed QA gates

  • Deploy: layers of agentic review happen, with human review reserved for regulated or critical code, enforced through hooks as approval gates

  • Maintain: agents monitor live deployments and diagnose issues, writing fixes back into the loop as new intent.md files

Most organizations sit somewhere between these two ends, not fully at either one. The important part isn't that AI replaces humans at every stage. Humans remain accountable for every decision that requires judgment. What changes is where that judgment gets applied, and how much of the repetitive, structured work around it gets handed off.

What Spec-Kit and SDD Actually Are

    Spec Driven Development (SDD) flips the usual order of operations. Instead of writing code first and documentation later (if at all), you write a clear, structured specification of what you want built, and that spec becomes the thing both humans and AI agents act on.
    Spec-Kit is the tooling layer for this. It gives you a structured way to capture intent, requirements, and constraints in a format an AI agent can actually read and act on, not just a document that sits in a wiki nobody opens again.
    Test Driven Development (TDD) pairs naturally with this. If the AI is going to generate code from a spec, having tests defined up front gives you a concrete, automatable way to check whether what got generated actually does what the spec said it should.

Why Human Review Still Matters

    None of this removes the need for human review, it changes where it's applied. The playbook framing, layers of agentic review with human review reserved for regulated and critical code, matches what I've found in practice. AI-generated code can look complete and still miss intent, especially on anything with real business logic or edge cases that weren't spelled out clearly in the spec. The spec is only as good as what you put into it, and catching gaps in the spec itself is still a human job.

What's Next

    In Part 2, I'll go through specific "plays" from the playbook, the concrete starting points for adopting this stage by stage, and how they map to what I'm actually doing day to day. After that, I'll get into KARIMO, the AI-native coding system I forked and have been personalizing, and KARIMO-Reviewer, the automated PR review agent I built alongside it.

Comments

Post a Comment

Popular posts from this blog

Building a Local RAG System with Ollama and Gemma: A Complete Guide - Part 1

Image
Building a Local RAG System with Ollama and Gemma: A Complete Guide Retrieval-Augmented Generation (RAG) has revolutionized how we interact with large language models by combining the power of information retrieval with text generation. In this comprehensive guide, we’ll walk through creating a complete RAG system that runs entirely on your local machine using Ollama and the Gemma 2B model. Why Build a Local RAG System? Before diving into the implementation, let’s understand why building a local RAG system is beneficial: Data Privacy : Your sensitive documents never leave your machine Cost Efficiency : No API costs or usage limits Offline Capability : Works without internet connectivity Customization : Full control over the model and parameters Scalability : Process large document collections without external constraints What is RAG? RAG (Retrieval-Augmented Generation) combines two key components: Retrieval System : Searches for relevant...
Read more

Building a Local RAG System with Ollama and Gemma: A Complete Guide - Part 2

Image
Building a Local RAG System with Chat Memory Building a Local RAG System with Chat Memory Using Redis This is a continuation of our comprehensive guide on building a local RAG system with Ollama and Gemma. If you haven’t read Part 1, we recommend starting there to understand the foundational concepts and basic implementation. In our previous article, we successfully built a functional RAG (Retrieval-Augmented Generation) system that could process documents and answer questions based on their content. However, our system had one significant limitation: it couldn’t remember previous conversations or maintain context across multiple interactions. Today, we’ll enhance our RAG system by adding chat memory functionality using Redis, enabling it to maintain conversation history and provide more contextual responses. This upgrade transforms our stateless question-answering system into an intelligent conversational AI that can reference previous exchanges. Why Add Ch...
Read more

Building a Local RAG System with Ollama and Gemma: A Complete Guide - Part 3

Image
Deploying Your Local RAG System with Chat Memory to Google Cloud Platform Deploying Your Local RAG System with Chat Memory to Google Cloud Platform This is the third installment in our comprehensive series on building and deploying RAG (Retrieval-Augmented Generation) systems. In Part 1 , we built a foundational RAG system using Ollama and Gemma. In Part 2 , we enhanced it with Redis-based chat memory functionality. Now, we’ll take the next crucial step: deploying our memory-enhanced RAG system to Google Cloud Platform (GCP) for production use. Moving from local development to cloud deployment opens up new possibilities for your RAG system. You’ll gain better accessibility, scalability, and the ability to serve multiple users simultaneously while maintaining the same powerful local AI capabilities we’ve built. Why Deploy to Google Cloud Platform? Before diving into the deployment process, let’s understand why GCP is an excellent choice for hosting your RAG ...
Read more