AI Tools I Use

My daily AI toolkit for development and productivity

Updated March 2026

I use AI tools daily to accelerate development, debug complex issues, and explore new approaches to problems. Here's my current toolkit—what I use, why I use it, and how each tool fits into my workflow.

🎯

Claude Code

Primary

My go-to for complex coding tasks, architecture decisions, and code review. Claude excels at understanding context and providing thoughtful, well-reasoned solutions.

I use it for:

  • Complex refactoring and architecture planning
  • Code review and security analysis
  • Writing documentation and technical specs
  • Debugging tricky issues
claude.ai →
💫

Qwen

Versatile

Excellent for quick iterations, code generation, and when I need a second opinion on Claude's suggestions. Great balance of speed and accuracy.

I use it for:

  • Quick code snippets and boilerplate
  • Alternative perspectives on solutions
  • Multi-language support
  • Fast prototyping
qwen.ai →
🌪️

Mistral

Fast

Lightweight and fast. Perfect for quick questions, simple transformations, and when I need answers without waiting.

I use it for:

  • Quick syntax questions
  • Simple code transformations
  • Text summarization
  • Fast iterations on small tasks
mistral.ai →
✨

Gemini

Research

Strong research capabilities and Google ecosystem integration. I use it when I need to verify information or explore new topics.

I use it for:

  • Research and fact-checking
  • Google ecosystem integrations
  • Multi-modal tasks (images + text)
  • Exploring new technologies
gemini.google.com →
🦙

Ollama

Local

Run AI models locally on my machine. Essential for privacy-sensitive work, offline development, and experimenting with different model sizes.

I use it for:

  • Privacy-sensitive code review
  • Offline development
  • Testing different model sizes (7B, 13B, 70B)
  • Custom fine-tuned models
ollama.ai →

How I Use These Tools Together

1

Start with Claude Code

For complex problems, I begin with Claude Code. Its deep reasoning helps me understand the problem space and explore architectural options.

2

Verify with Qwen or Gemini

I cross-reference Claude's suggestions with Qwen for code-heavy tasks or Gemini for research-heavy topics. Different models catch different issues.

3

Quick iterations with Mistral

For small refinements, syntax questions, or quick transformations, Mistral is fast and doesn't waste tokens.

4

Local testing with Ollama

Before deploying AI-generated code to production, I run it through local models for a final privacy check and to catch any remaining issues.

Why Use Multiple AI Tools?

🎯 Different Strengths

Each model excels at different tasks. Claude for reasoning, Qwen for code, Mistral for speed, Gemini for research, Ollama for privacy.

✅ Cross-Verification

Multiple models catching the same issue gives me confidence. When they disagree, it reveals edge cases I should consider.

🔒 Privacy & Control

Local models via Ollama let me work with sensitive code without sending it to external APIs.

⚡ Speed vs. Depth

Sometimes I need a quick answer (Mistral), sometimes I need deep analysis (Claude). Having both options is essential.

Questions About My AI Workflow?

I'm always happy to discuss AI tools, workflows, and how I integrate them into development.

Get In Touch Read My Blog