Practical field notes on the parts of AI systems that have to survive production: agents, evaluation, security, memory, retrieval, model engineering, and infrastructure.
Choose a route by problem below: agent systems, evaluation, or retrieval and language. For a shorter decision guide, use Fast Answers.
Start here
Agent systems
A connected path through reasoning, memory, tools, security, runtime, and harness engineering.
Evaluation
Evaluate agent behavior and every stage of a retrieval-augmented generation system.
Retrieval & language
Build ranking pipelines, choose document AI systems, and structure language reliably.
Flagship series
Swipe through all six parts →
Developer Tooling · AI Agents
Map pydantic to Zod, FastAPI to Hono, and Celery to a Postgres queue, then build a streaming agent service in a pnpm TypeScript monorepo.
· 23 min
AI Agents Part 6
How the control code around an AI agent authorizes tools, preserves progress, verifies completion, and earns each added component through ablation.
· 24 min
Model Engineering · AI Infrastructure
A practical guide to choosing LLM and diffusion-model quantization paths by bottleneck, hardware, runtime kernel, calibration data, and deploy checks.
· 23 min
AI Agents
Why production AI agents need schema-guided memory with temporal validity, provenance, and structured conflict handling instead of raw text vector recall.
· 14 min
AI Agents
How to evaluate production AI agents by turning traces into versioned regression datasets, trajectory metrics, calibrated judges, and CI gates.
· 21 min
AI Agents Part 5
How to run long-running AI agents in production with sessions, harnesses, sandboxes, checkpoints, traces, and deployment patterns that recover cleanly.
· 37 min
Retrieval and Language Systems
RAG evaluation metrics for parsing, retrieval, reranking, generation, citations, and production telemetry, with code and failure modes.
· 35 min
AI Agents Part 4
Why LLM guardrails are not enough for AI agent security, and where permissions, sandboxes, HITL, MCP scoping, and policy checks fit.
· 35 min
Retrieval and Language Systems
NER in 2026 means choosing between GLiNER, spaCy, Transformers, and LLM extraction for latency, accuracy, and schema control.
· 30 min
AI Agents Part 3
Compare JSON tool calling, MCP, Skills, CLI, and code execution for AI agents, with ACI design rules and the production trade-offs that matter.
· 19 min
Model Engineering · AI Infrastructure
A practitioner's reference to 45 LLM engineering concepts for production systems, spanning inference, training, architecture, deployment, and operations.
· 60 min
Retrieval and Language Systems
OCR in 2026 means choosing between classical pipelines and VLMs for text, layout, tables, and document extraction.
· 12 min
AI Infrastructure
Benchmarks comparing Polars, DataFusion, Daft, Ray Data, Pandas, and Spark on tabular and multimodal workloads, with code and decision rules.
· 14 min
AI Agents Part 2
How to design AI agent memory with checkpoints, PostgreSQL or Redis, vector stores such as Qdrant, and file-based memory for long-running systems.
· 18 min
Retrieval and Language Systems
How to build a search ranking stack with BM25, dense embeddings, hybrid RRF, cross-encoder reranking, and LLM listwise reranking on Amazon ESCI.
· 24 min
AI Agents Part 1
Compare ReAct, ReWOO, and Plan-and-Execute for AI agents. LangGraph examples show how each loop trades adaptability for cost, latency, and planning structure.
· 12 min
Model Engineering
A technical walkthrough of DeepSeek's Manifold-Constrained Hyper-Connections (mHC), residual stream width scaling, Sinkhorn routing, and training stability.
· 9 min
AI Agents
What public Enterprise RAG Challenge 3 submissions reveal about prompt iteration, validation, context strategy, tool wrappers, and agent architecture.
· 9 min
Model Engineering
When to fine-tune LLMs, when to use RAG or prompting, and how LoRA, QLoRA, DoRA, Unsloth, Axolotl, datasets, evals, and deployment fit.
· 19 min
AI Agents · Model Engineering
How Schema-Guided Reasoning uses vLLM, XGrammar, Pydantic schemas, and constrained decoding to enforce structurally valid LLM outputs.
· 10 min
Model Engineering · AI Infrastructure
Evaluate and deploy LoRAX for long-tail LoRA serving, with current APIs, Kubernetes chart hardening, adapter security, cache-aware routing, and vLLM trade-offs.
· 10 min
AI Agents
How domain-driven design helps AI agents use shared language, bounded contexts, services, repositories, and deterministic business rules.
· 8 min
AI Agents
How to design context engineering for AI agents: context windows, instruction hierarchy, retrieval, memory, tools, guardrails, provenance, and compression.
· 14 min
AI Agents · Developer Tooling
Build a custom MCP server with uv and FastMCP, expose ML feature-store tools, test them locally, and connect the server to Claude Desktop.
· 8 min
Model Engineering
How to read open-weight model names by separating training role, architecture, quantization method, storage format, runtime, and hardware fit.
· 9 min
Developer Tooling
Compare local image-generation tools on macOS by setup, model support, workflow control, reproducibility, extensions, and maintenance status.
· 6 min
Model Engineering · Developer Tooling
Compare local LLM tools on macOS by interface, model format, memory headroom, API exposure, and Apple Silicon development workflow.
· 6 min
Developer Tooling
How pyproject.toml works for build systems, project metadata, runtime dependencies, development groups, CLI entry points, and tool configuration.
· 5 min
Developer Tooling
Zsh startup files explained: when ~/.zprofile and ~/.zshrc load on macOS and Linux, what belongs in each file, and how to diagnose the running shell mode.
· 4 min
AI Infrastructure
What foundation models changed in MLOps: release units, retrieval, tools, evaluation, tracing, serving, and operational risk.
· 7 min
Model Engineering · AI Infrastructure
How to scale large language models across GPUs and nodes with data, fully sharded, tensor, pipeline, context, and expert parallelism.
· 8 min
Developer Tooling
A MacBook setup checklist for AI engineering: Xcode tools, Homebrew, Python with uv, Docker, terminal setup, and VS Code.
· 6 min
Developer Tooling
How I use uv on macOS for Python installs, project dependencies, lock files, inline scripts, and one-off CLI tools.
· 6 min
Nothing here yet.
