DEV Community

Amariah Kamau profile picture

Amariah Kamau

Certified Data Engineer (AWS & GCP) and Full Stack Developer with specialized expertise in Agentic Workflows and Scalable Systems.

Pronouns

He/Him

The harness, not the model: how to make a weak or local model reliable enough to ship

The harness, not the model: how to make a weak or local model reliable enough to ship

1
Comments 4
7 min read

Want to connect with Amariah Kamau?

Create an account to connect with Amariah Kamau. You can also sign in below to proceed if you already have an account.

Already have an account? Sign in
I shipped two PRs into Alibaba's qwen-code using only open-weight models. Here's the honest version.

I shipped two PRs into Alibaba's qwen-code using only open-weight models. Here's the honest version.

Comments 3
5 min read
Atlarix vs opencode on Terminal-Bench 2.0 — same model, only the harness changes (k=1, receipts included)

Atlarix vs opencode on Terminal-Bench 2.0 — same model, only the harness changes (k=1, receipts included)

Comments
4 min read
Show dev: I built an AI agent workstation in Nairobi for open-weight and local models

Show dev: I built an AI agent workstation in Nairobi for open-weight and local models

2
Comments
3 min read
We Tried to Reduce LLM Context Usage in a Multi-Repo Codebase. The AI Used More Tokens, Not Less. Here's Why That's Correct.

We Tried to Reduce LLM Context Usage in a Multi-Repo Codebase. The AI Used More Tokens, Not Less. Here's Why That's Correct.

Comments
5 min read
Build an AI Agent That Actually Understands Your Codebase (Without Switching Editors)

Build an AI Agent That Actually Understands Your Codebase (Without Switching Editors)

Comments
7 min read
How I reduced AI codebase context from 100K to 5K tokens using a graph-based RAG — and why I deleted it

How I reduced AI codebase context from 100K to 5K tokens using a graph-based RAG — and why I deleted it

Comments 1
3 min read
What two hackathons taught me about agent architecture — and how it's reshaping Atlarix

What two hackathons taught me about agent architecture — and how it's reshaping Atlarix

Comments
2 min read
The Private Sovereign AI Workstation — Why I built Atlarix differently

The Private Sovereign AI Workstation — Why I built Atlarix differently

Comments
3 min read
Atlarix v3.7 is live — now Apple Notarized

Atlarix v3.7 is live — now Apple Notarized

Comments
1 min read
I Got Tired of Re-Explaining My Codebase to AI Every Single Session

I Got Tired of Re-Explaining My Codebase to AI Every Single Session

1
Comments
6 min read
Cursor vs Windsurf vs Atlarix: Which AI Coding Assistant Should You Choose in 2026?

Cursor vs Windsurf vs Atlarix: Which AI Coding Assistant Should You Choose in 2026?

Comments
4 min read
Why I Built a Local-First AI Coding Agent Instead of Another Chat Assistant

Why I Built a Local-First AI Coding Agent Instead of Another Chat Assistant

Comments
2 min read
🟩 commit-checker v0.6.1 – Track GitHub streaks, TIL logs, and unlock CLI achievements

🟩 commit-checker v0.6.1 – Track GitHub streaks, TIL logs, and unlock CLI achievements

Comments
2 min read
loading...