/blog · playbook notes

Notes from the workshop floor.

Slow, careful posts on what we've learned shipping AI-native apps.

A practical RAG chunking playbook

Chunking is the single highest-leverage knob in a RAG system. Here is how we choose chunk size, overlap, and structure for real corpora.

May 30, 2026Read

5 min

Stateful Agents, Standardized Protocols, and Native UI: The Next Era of Developer Tooling

The developer landscape is undergoing a major transition. Engineers are moving past brittle, stateless LLM wrappers toward resilient, state-managing agentic systems on the backend, balanced by a return to high-performance, lightweight craftsmanship on the frontend. The current focus centers on building reliable developer tooling, predictable AI execution, and hardware-accelerated user interfaces.

May 30, 2026Read

5 min

Shipping AI-Native Apps on a Two-Week Cycle

How we structure a KiteLabs build: week one is a vertical slice, week two is the eval flywheel.

May 30, 2026Read

7 min

Evals That Actually Catch Regressions

Why golden datasets, LLM-as-judge, and production traces are three different evals — and you need all three.

May 30, 2026Read

6 min

RAG Is Not a Vector Database

Retrieval quality lives in chunking, query rewriting, and rerankers — not in which vector store you picked.

May 30, 2026Read

8 min

Designing Planner–Executor Agents That Don't Drift

A pragmatic recipe for splitting reasoning from action so your agents stay on the rails in production.

May 30, 2026Read

/newsletter · low-volume

New posts, straight to you.

One email when something new ships. No spam, unsubscribe anytime.