Skip to content
§ Research Archive

Long-form notes on production AI systems.

An evolving archive of technical research, design decisions, and operational lessons from building AI infrastructure.

§ What this section is

The conceptual layer of the lab — the thinking that precedes every system shipped here.

Research is where loose questions get sharpened into operational hypotheses. Each area below is a long-running thread, not a topic page — open it to see what's currently being studied, what's been learned, and which build logs the thread feeds.

Start with the founding manifesto if you're new. Otherwise, follow any thread that maps to a problem you're actually facing. The archive is designed to be entered sideways.

Founding noteRES/0001Vol I · Issue 01~14 min

The Researcher’s Path: a founding note on building intelligence, in public, for the long term.

On fragmentation, the operator’s mindset, public iteration, and AI as leverage infrastructure — the thesis behind this lab.

AI Agents

Designing reliable multi-step agents without losing your mind

A practitioner's taxonomy of failure modes and the control structures that contain them.

May 04, 2026
18 min
RAG Systems

Beyond top-k: retrieval architectures for production knowledge work

Hybrid retrieval, reranking, and the operational tradeoffs nobody publishes benchmarks for.

Apr 22, 2026
24 min
Evaluation

Eval pipelines as an engineering discipline, not a checkbox

Treating evals like CI: traces, fixtures, regressions, and budgets.

Apr 09, 2026
15 min
Memory Systems

A taxonomy of agent memory: working, episodic, semantic, procedural

Mapping cognitive science abstractions onto durable agent state.

Mar 27, 2026
21 min
Orchestration

Graph orchestration vs. handcrafted state machines

Where general orchestrators win, and where they collapse under real workloads.

Mar 14, 2026
19 min
Reliability

Observability for non-deterministic systems

Trace shapes, sampling strategies, and what 'p99' even means for an LLM call graph.

Feb 28, 2026
22 min
Deployment

From notebook to service: a minimal production envelope

The smallest defensible deployment surface for AI features.

Feb 12, 2026
16 min
AI Agents

Tool-use as an interface design problem

Why most tool-using agents fail at the API boundary, not the model.

Jan 30, 2026
13 min