The Tech Trek by Tech Chick👑👩🏻‍💻 - Medium

Not Every AI Agent Needs RAG. Sometimes a Markdown Folder Is Enough.

Chandana — Thu, 12 Nov 2026 22:27:43 GMT

We’ve become so used to the concept of adding embeddings and vector databases to AI applications that we rarely stop to ask whether the…

Chandana — Fri, 06 Nov 2026 07:44:21 GMT

Reliability is tougher to measure than task completion. Here’s how to evaluate what actually makes a Claude Code harness better.

Chandana — Mon, 02 Nov 2026 20:23:52 GMT

Here’s what it means and how China, the US, and India are feeling it

Chandana — Wed, 28 Oct 2026 20:49:39 GMT

Getting AI to work in a demo and making it deliver real value across an organisation are two very different problems.

Chandana — Tue, 20 Oct 2026 09:18:45 GMT

How tracing, observability, and evaluations help you understand failures and build more reliable AI agents.

Chandana — Fri, 16 Oct 2026 20:18:25 GMT

How context management, guardrails, verification, and recovery turn a working AI agent into one you can actually trust.

Chandana — Sat, 10 Oct 2026 09:06:04 GMT

What harness engineering is all about and how execution loops, tools, and feedback turn a language model into an agent that actually gets…

Chandana — Tue, 29 Sep 2026 19:52:02 GMT

What I learned at a MarTech meetup about using AI for content, design, SEO, leads, email and marketing automation

Chandana — Tue, 22 Sep 2026 20:47:31 GMT

Why LLM-as-a-Judge may not need another LLM, and how Jev uses probabilistic decisions to make AI evaluation faster, cheaper, and less flaky

Chandana — Wed, 16 Sep 2026 19:12:11 GMT

From Agent Failures and Retries to LangGraph and GraphRAG: Building Smarter AI Agent Workflows