Stop paying for every LLM: My $0 AI toolkit
From local LM Studio to cloud-based Hermes: How I keep building when Claude Code credits run out.
AI Weekender
I write weekly for data scientists and engineers building with AI tools and coding agents.
Live Last read · last published · next check
Latest episodes
Get Hermes + Telegram on Your $0 Cloud VM (Part 2)
Step-by-step guide on setting up Hermes + Telegram on your cloud VM from part 1.
Set Up a $0 Cloud VM to Run Your Hermes Agent (Part 1)
How to get a free-tier cloud VM you can SSH into and leave running 24/7.
I Built a $0 Hermes Agent I Text From Telegram
Honest look at my always-on Hermes on a $0 cloud virtual machine + NVIDIA API free tier
My $0 Hermes Agent Setup
I wanted to try Hermes without paying for Portal, so I pointed the agent at NVIDIA's free API tier instead.
Prompt vs RAG vs Fine-Tuning: Which Fix Do You Need?
Diagnose knowledge vs behavior failures, then choose prompting, retrieval, or training
How I Fine-tuned Qwen3-4B on a Single GPU in 5 Steps
0% → 83% correctly classified banking support messages: fine-tune a small LLM to classify customer intent.
Glossary of Fine-Tuning Methods Explained
SFT, LoRA, QLoRA, DPO, RLHF & GRPO: what each acronym means and how to pick the right fine-tuning method from the labels you already have.
Fine-Tuning LLMs: When to Do It and When Not To
When LLM fine-tuning beats prompting and RAG and when to skip it
When Your Coding Agent Says Done (And Isn’t)
Session fixes for false progress, doom loops, and instruction drift in Claude Code and Cursor (after re-prompting made my stuck sessions worse).
How Coding Agents Actually Work Under the Hood (and Why They Go Wrong)
When the agent says it made a fix but your app still breaks: a clear read on Cursor, Claude Code, and what the session is really doing.
6 LLM Prompting Techniques for Data Scientists and Engineers in 2026
Six techniques matched to six failure modes, including inconsistent output formats, shallow reasoning, instruction drift, and more.
Should You Run an LLM on Your Phone?
What works today with Google AI Edge Gallery, what breaks first on real devices, and how to pick a model your phone can actually handle
Before You Upgrade Embeddings, Check Your Query Logs
What I learned at PyCon and from my own RAG failures
How to Learn AI by Building Real Projects (Not Tutorials)
A practical system: ship a live MVP, iterate from user feedback, and learn only what blocks you
From Indexing to Answering: A RAG Pipeline You Can Click Through
Interactive explainer on the query loop, ingestion vs retrieval, configs, deploy for my AI assistant I'm presenting at PyCon Austria.
Claude Code Keeps Forgetting Your Stack. These 5 Files Fix That.
Paste-ready CLAUDE.md template, settings.json deny list with reasoning, path-scoped UI rules, two custom /project: commands
8 Ways Vibe-Coded Apps Leak Data
Common AI-generated code failure modes (authorization, CSRF, prompt injection) and the paste-ready agent rules to guard against all of them.
16 Claude Code Commands That Actually Changed My Workflow
Sixteen Claude Code slash commands for the model context window, planning, diffs, and spend
How to Pick Your Next AI Project (Before You Waste a Weekend)
I've built 10+ AI projects. Only 2 are still running. Here's what I'd do differently.