Blog

Short notes on shipping generative AI

Thirty short articles on shipping generative AI in production: RAG quality in production, choosing an LLM in 2026 across GPT, Claude, Gemini, and open-source, daily Claude Code setup with Cursor and MCP, coding-agent workflow, LLM evaluation, AI memory for agents, field lessons before deploying a first agent, deployment-at-scale lessons from Google and Microsoft, AI agents in 2026, production-first experiments after POCs, vibe coding, LLM deployment, production RAG, data quality, business scoping, AI audit, architecture, AI ROI measurement, prompt engineering, team integration, common failure modes, evals, agents, AI coding agents, multi-agent systems, automation, and practical framework choices such as LangGraph versus CrewAI once the demo is over.

27 August 2026

What a GenAI project costs in 2026: budget, timeline, ROI

A practical 2026 guide to GenAI project cost, timeline, ROI, scoping, POCs, production delivery, run costs, and Build vs Audit vs Train decisions.

22 August 2026

RAG in production: the 7 mistakes that break quality (and how to fix them)

The 7 RAG mistakes I fix in production: chunking, retrieval evals, reranking, metadata filters, hallucination guardrails, latency, cost, and monitoring.

18 August 2026

Choosing your LLM in 2026: GPT, Claude, Gemini, open-source — how I decide

My practical framework for choosing GPT, Claude, Gemini or open-source models in production: task fit, latency, cost, evals, context and EU constraints.

29 July 2026

Coding with an AI agent: my daily Claude Code + MCP setup

My daily setup for coding with Claude Code, Cursor, and MCP servers: repo context, eval rails, review, deploy discipline, and Brut/Ellipse production lessons.

27 July 2026

Claude Code in production: my AI engineer workflow

My concrete workflow with Claude Code, Cursor, and MCP for shipping client AI builds: scoping, context, coding agents, reviews, evals, and team adoption.

22 July 2026

Evaluating LLMs in production: building an eval suite that actually matters

How to build useful LLM evals in production: task success, groundedness, latency, cost, golden datasets, LLM judges, and CI.

30 June 2026

RAG in production: the 5 mistakes I keep seeing everywhere

The five RAG mistakes I keep seeing in production: poor chunking, no reranking, no evaluation loop, treating RAG as a silver bullet, and skipping latency and cost budgets — with concrete fixes.

29 June 2026

Why RAG Is No Longer Enough: Toward Intelligent Memory Systems

RAG is still useful, but production agents need real memory: short-term state, episodes, semantic knowledge, procedures, and controlled update rules.

27 June 2026

What I wish I'd known before deploying my first production agent

A candid production-agent retrospective: prompt drift, context-window cost, fallback logic, monitoring, UX, and the user trust work beyond the model.

27 June 2026

What Google and Microsoft taught me about deploying at scale

Personal lessons from Google and Microsoft applied to AI deployments: reliability, observability, documentation, speed, and experimentation for real LLM systems.

24 June 2026

AI agents in 2026: what changed (and what hasn't)

A field report on AI agents in production in 2026: what is genuinely better, what is still hard, and which 2022-2023 principles still hold.

22 June 2026

Why I stopped doing POCs — and what I do instead

AI POCs often create the illusion of progress. What turns an idea into a useful product is a smaller, real-world experiment designed for production from day one.

21 June 2026

Vibe coding: what it actually changes for product teams

Vibe coding is not a magic shortcut. It is a faster way to turn product judgment into working software, as long as humans still own scope, architecture, quality, and responsibility.

19 June 2026

What I learned deploying LLMs for real clients

Honest lessons from LLM deployment with real clients: what breaks between demo and production, how latency, prompt engineering, cost, adoption, and ownership decide whether AI in production lasts.

18 June 2026

AI and data: why data quality is the real bottleneck

Why most AI projects fail less because of the model than because of data quality: unlabeled data, inconsistent schemas, missing context, outdated exports, and the practical checklist before starting production AI.

29 May 2026

Measuring the ROI of an AI project: what leadership actually needs

How to measure AI ROI without falling for vanity metrics: the KPIs leadership actually cares about, the baseline to define before the AI project starts, and the simple value story that survives the board room.

28 May 2026

Prompt engineering in 2026: what actually matters

A candid production take on prompt engineering in 2026: what still matters, what breaks as soon as the model changes, and what has actually become more important with structured output, tools, and evals.

27 May 2026

From AI audit to deployment: anatomy of a real project

What a serious AI project really looks like: discovery, honest AI audit, architecture choices, iterative build, team training, and production deployment.

26 May 2026

Building a multi-agent system: what I actually learned

A candid production account of building a multi-agent system: the orchestrator matters, contracts between AI agents matter even more, and state complexity arrives faster than most tutorials admit.

25 May 2026

Claude Code in production: lessons from the field

A candid field report on Claude Code in production: excellent on bounded execution and refactors, much weaker whenever context, priorities, or guardrails stay implicit.

24 May 2026

When to automate and when not to: the real trade-off

In production, the real question is not whether AI automation is possible, but whether it reduces the cost of work without destroying useful human judgment.

23 May 2026

Corporate AI training: what actually works

Corporate AI training becomes useful when it starts from real workflows, gives people tools they can use the next day, and turns a few participants into durable internal champions.

22 May 2026

LangGraph vs CrewAI: What I Learned in Production

Two frameworks, two philosophies. After shipping agents with both tools, here is what I actually learned.

22 May 2026

Why a standalone AI agent isn't enough: the role of business context

The real differentiator for a production AI agent is rarely the model or framework. It is the quality of the business context shaping its decisions, constraints, and escalation paths.

20 May 2026

Why LLM evaluation is the real engineering work

Why the real leverage on a production LLM system is rarely the prompt or the model, but an evaluation loop that catches regressions before users do.

18 May 2026

MCP, LangGraph, agents: what real production projects actually teach you

What production agent systems teach in practice about MCP, LangGraph, multi-agent coordination, and the guardrails that prevent a slick demo from becoming an operational mess.

15 May 2026

How to evaluate an AI model in production: metrics, evals, and pitfalls to avoid

A practical framework for evaluating an LLM system in production without confusing benchmark scores with real quality, reliability, or business value.

12 May 2026

Generative AI in Business: Where to Actually Start

A practical framework for starting a generative AI initiative in a business without getting trapped in noise, demos, or bloated roadmaps.

11 May 2026

Why your AI proof of concept never makes it to production (and how to fix it)

Why AI proof-of-concept projects stall before production, and how to fix the data, ops, governance, and technical debt issues blocking rollout.

10 May 2026

How to integrate generative AI into a tech team in 2025

A practical framework for integrating generative AI into a tech team in 2025 without sacrificing quality, security, or accountability.

3 May 2026

The 3 most common mistakes in enterprise AI projects

Three recurring mistakes that slow down enterprise AI projects, and a more durable way to scope, build, and ship them.