Build logs, guides, and field notes
What shipped, what broke, and what it really cost — from running production AI systems.
Latest articlesWhat shipped, what broke, and what it really cost — from running production AI systems.
Gumloop Review
Gumloop reviewed for lead research, scraping, documents and email: a step-by-step setup, current credits, approval controls and alternatives.Oct 22, 2026Build
Meta Muse Privacy: Settings and a Workplace Policy
What Meta's Muse keeps, trains on and shares, with verified privacy controls, reported incidents and a short staff policy.Oct 22, 2026AI
Synthetic Data Generation: LLM Training and Test Data
Generate LLM evals, fine-tuning examples and test fixtures with a worked recipe, filtering, deduplication, human checks and provider terms.
Oct 22, 2026Build
RLHF Explained: Human Feedback, DPO and Fine-Tuning
How RLHF turns human rankings into model behavior, how DPO and verifiable rewards differ, and which preference fine-tuning APIs a small company can use.Oct 22, 2026Build
Model Router: How LLM Routing Works and How to Test It
Learn how LLM model routing works, compare documented options, and test quality with a support-app example, cost formula, and held-out eval set.Oct 21, 2026Build
Semantic Caching: How It Works and When to Reuse Answers
Understand semantic caching, its four-step lookup, and how thresholds, expiry and user scope prevent similar questions from receiving the wrong answer.Oct 21, 2026Build
Human in the Loop AI: Approvals, Review Queues and Escalation
Decide when AI agents need approval, sampled review or escalation, with a support-agent example, current tool docs and practical logging rules.Oct 21, 2026AI
Pydantic AI vs LangChain
Choose a Python agent framework with matching ticket-triage code, validated outputs, LangGraph workflows, tracing, evals and migration guidance.Oct 21, 2026Build
Zed vs Cursor 2026: Editing Feel, Agents and Teamwork
Zed vs Cursor for developers: compare editing feel, agents, model choice, collaboration and the limits of switching or using both.Oct 21, 2026Build
DeepSeek vs Qwen
Choose DeepSeek or Qwen for coding, vision, long context and self-hosting. Current models compared by job, with verified licenses and hosting scopes.Oct 21, 2026AIThe lanes.Pick the one you came for.
243 pieces
AI
Models, agents and coding assistants, image, video and voice tools: reviewed, priced and compared.297 piecesBuild
Ship software with AI, whether or not you write code: coding agents, app builders, automation, MCP and the infra under it.123 piecesDesign
AI as a craft tool for designers, brand owners and creative directors: image, video, type, and the prompts behind the work.202 piecesBusiness
Voice and phone agents, customer service, sales and marketing, operations and AI search visibility, for the people who run a business.8 piecesPlaybooks
AI for your niche: the workflows that actually move, tool by tool, priced52 piecesExplained
What shipped, explained: every real release, decoded for the people who use it0 piecesLab
The agent journey, entry by entry: what shipped, what broke, what was decided, in the ventures I run924–925 / 925
Claude for Small Business: What It Kills in a $50K Stack
On May 13 Anthropic launched Claude for Small Business, a Cowork add-on with eight native connectors and fifteen ready-to-run agentic workflows priced at zero incremental cost on top of an existing Claude license. This piece is the…May 14, 2026AI
Cloudflare's '100x engineer', a 6-agent operator's take
Matthew Prince cited 2x-100x AI productivity gains to justify 1,100 layoffs. I run 6 production agents on the exact stack he's describing. The 100x engineer framing is conceptually wrong, and founders copying it into H2 headcount plans are about to cut the wrong people.May 12, 2026BusinessPrefer this site in Google
Add as a preferred sourceon GoogleMark omidsaffari.com as preferred and Google lifts it in Top Stories, AI Overviews and AI Mode for you.
Newsletter
One letter, every Sunday.Working systems, not hot takes.
Weekly. No spam. Unsubscribe anytime.