PLAYBOOKS · DATA · FIELD NOTES

Notes from the engineering side

Long-form essays from current engagements. No AI hype. No agency template.

Virtual CTO Services: When You Need One and How to Hire
Featured

Virtual CTO Services: When You Need One and How to Hire

A virtual CTO gives you executive-grade engineering leadership without the cost of a full hire. Here is when to bring one in, what a real engagement looks like, and how to pick the right one.

Apr 28, 2026 · 6 min read
Browse all topics
Block, allow, or charge: the new controls for AI crawlers
AI

Block, allow, or charge: the new controls for AI crawlers

For thirty years the web ran on one deal: let crawlers in, get referrals back. AI crawlers took the content and stopped sending the traffic. Cloudflare's answer turns a binary — block or allow — into a third option: charge. Here is what the controls actually are, when to reach for each, and why 'block' is a policy, not a guarantee.

August 1, 2026 · 9 min read
Notion Mail and the Cost of Building What Nobody Asked For
Product

Notion Mail and the Cost of Building What Nobody Asked For

Notion is shutting Notion Mail 16 months after launch because more than half its users never opened the inbox. The shutdown is the cheap part. The quarter of engineering that built a product the market never asked for is the loss — and the signals that would have caught it were checkable before the build.

July 31, 2026 · 8 min read
The Biggest AI Checks This Week Went to Running Models, Not Building Them
AI

The Biggest AI Checks This Week Went to Running Models, Not Building Them

One week of AI funding rounds — Baseten, Groq, Upscale AI, General Intuition, AppsFlyer — read as a buy-vs-build map for a CTO. Sort the news by whether your team would ever touch it, and the week tells you which layers just became infrastructure to rent.

July 30, 2026 · 9 min read
"The AI Did It" Is Not a Defense
AI

"The AI Did It" Is Not a Defense

The fear is that AI creates an accountability-free zone — that anyone can hide behind the model when it's wrong. Courts and professional bodies are ruling the opposite way. The liability lands on whoever deployed the AI, and that's the filter separating a specialist from a prompter.

July 29, 2026 · 8 min read
How Anthropic builds engineering teams
Leadership

How Anthropic builds engineering teams

The numbers everyone quotes from Anthropic — 80% of merged code AI-generated, teams running four projects at once, more product managers not fewer — are real. They're also the output of something most teams don't have: senior-only hiring and review infrastructure that trusts the volume. Copy the org chart without the substrate and you get a smaller multiplier plus a pile of unreviewed code.

July 28, 2026 · 8 min read
AI coding agents are the default. Now the bottleneck is you.
AI

AI coding agents are the default. Now the bottleneck is you.

Adoption is settled, so the interesting question is what broke. Generation scaled faster than verification did: AI-assisted PRs merge at 32.7% against 84.5% for human ones, and the delay is getting picked up, not the reviewing. What the second-order shifts are, and the four numbers to instrument this quarter.

July 27, 2026 · 23 min read
The Cyber Resilience Act is product-safety law for software. Two clocks are already running.
Security

The Cyber Resilience Act is product-safety law for software. Two clocks are already running.

The CRA points EU product-safety law at software. Ship an installable artifact — or build bespoke, or white-label someone else's code — and you're a manufacturer, with a 24-hour reporting duty live from September 2026 that reaches everything you have already shipped.

July 26, 2026 · 21 min read
Model Size Is a Deployment Axis
AI

Model Size Is a Deployment Axis

Large and small language models aren't competitors ranked on one axis — they're different tools. Match the class to the task's shape and 'which is better' becomes a routing table. A per-use-case verdict, with the benchmark numbers, including the ones that expired.

July 25, 2026 · 19 min read
AI didn't replace engineers. It moved the demand to verification.
AI

AI didn't replace engineers. It moved the demand to verification.

Engineering held up better than any function in tech while new-grad hiring fell 65% — one dataset, both facts. AI didn't cut demand for engineers; it moved the demand from producing code to verifying it, and verification is senior work. Why the reassuring aggregate is composite, and why the attribution is still contested.

July 24, 2026 · 24 min read
How to Measure AI's Impact on Engineering the Way Uber Does
AI

How to Measure AI's Impact on Engineering the Way Uber Does

Uber's 2026 AI budget was gone in a quarter, and none of their metrics could say what it bought. A field guide to the metrics that break at every era boundary — and the one tied to shipped outcomes that survives agents.

July 23, 2026 · 27 min read
Loops, Precisely: What to Automate, What to Queue, and What It Costs You
AI

Loops, Precisely: What to Automate, What to Queue, and What It Costs You

Agent loops pay only where the work recurs and a machine can verify it — otherwise you have built a queue, not a loop. The taxonomy of the five primitives, a flaky-test loop start to finish, and the review-capacity data behind the argument.

July 22, 2026 · 19 min read
Google Antigravity, Explored: A Map of the Whole Surface
AI

Google Antigravity, Explored: A Map of the Whole Surface

Google Antigravity is a broad, capable agentic IDE — but "100x engineering" is a framing, not a measurement, and almost every capability it ships already existed in Cursor, Claude Code, Devin, or Kiro. A capability-by-capability map of what is genuinely new, and the verification bottleneck it does not touch.

July 21, 2026 · 18 min read
The bottleneck moved to review. GitHub shipped a partial fix.
Engineering

The bottleneck moved to review. GitHub shipped a partial fix.

GitHub's Copilot reviewer now reads your AGENTS.md, so it reviews against the same house rules your coding agent follows. It clears the mechanical first pass — but the verification gap it's aimed at is the half it structurally can't reach.

July 20, 2026 · 8 min read
How OpenAI delivers low-latency voice
Engineering

How OpenAI delivers low-latency voice

Sub-second spoken replies don't come from a faster GPU. OpenAI's Realtime API gets there by deleting the handoffs in the old speech pipeline, streaming at every stage, moving turn-taking into the model, and pushing transport to the edge. A guide to the architecture — and the four moves that transfer to any real-time system you build.

July 19, 2026 · 9 min read
The model you build on is rented. Price the risk.
AI

The model you build on is rented. Price the risk.

Anthropic's two newest models went dark worldwide for 19 days on an export-control order — zero notice, and the provider didn't choose the timing either. Most teams price that risk at zero. Here is what it actually costs, and the seam that makes a swap a config change.

July 15, 2026 · 7 min read
Deploying Cloudflare One with agent skills: a working guide
Security

Deploying Cloudflare One with agent skills: a working guide

Cloudflare shipped its Cloudflare One deployment and migration know-how as agent skills plus an MCP server. How the pieces fit, how to adopt it without handing an agent your security control plane on day one, and where the human still owns the cutover.

July 14, 2026 · 7 min read
If you don't host it, you rent it: reading Odyssey's $1.45B against Fable's blackout
AI

If you don't host it, you rent it: reading Odyssey's $1.45B against Fable's blackout

Odyssey's $1.45B raise and Fable 5's overnight shutdown are the same story. How to architect for model-provider dependency: gateways, version pinning, and naming your no-fallback bets.

July 8, 2026 · 7 min read
Google's data agents and MCP Toolbox: what they're for and how to set one up
AI

Google's data agents and MCP Toolbox: what they're for and how to set one up

What Google's data agents and MCP Toolbox actually do, which databases you can connect, the realistic outcomes — and a hands-on setup: install, tools.yaml, and identity-scoped auth.

July 7, 2026 · 16 min read
GLM 5.2 against the shipping market: where the first frontier-adjacent open model wins, and where it doesn't
AI

GLM 5.2 against the shipping market: where the first frontier-adjacent open model wins, and where it doesn't

GLM 5.2 is the first open-weight model in the frontier coding conversation. Real benchmarks, cost-per-task math, and where it wins against Claude, GPT-5.5, and DeepSeek.

July 3, 2026 · 18 min read
Charge AI Bots for Content: AWS WAF Starts Billing Crawlers
AI

Charge AI Bots for Content: AWS WAF Starts Billing Crawlers

AWS WAF can now meter and charge AI bots per request instead of blocking them. What's GA, how pay-per-crawl works, and where the edge tollbooth breaks.

June 26, 2026 · 8 min read
AI Hiring Isn't Colder Than the Process It Replaced
AI

AI Hiring Isn't Colder Than the Process It Replaced

AI hiring feels impersonal — but high-volume human hiring already was. What a fair AI hiring process owes candidates: disclosure, a human path, bias audits.

June 25, 2026 · 8 min read
SOC 2 Compliance for SaaS Startups: When to Start, What It Costs, How to Pass
Security

SOC 2 Compliance for SaaS Startups: When to Start, What It Costs, How to Pass

SOC 2 for startups is the security framework US enterprise buyers ask for first. Here is when to start, the realistic cost, and a six-month plan that passes Type II without overbuilding.

June 23, 2026 · 7 min read
ISO 27001 for SaaS Companies: Process, Benefits and Cost
Security

ISO 27001 for SaaS Companies: Process, Benefits and Cost

ISO 27001 is the certification European enterprise buyers ask for and most US SaaS companies postpone. Here is the process, the realistic cost, and when it earns its place on the roadmap.

June 16, 2026 · 6 min read
Engineering Team Structure: Models That Scale from 10 to 100 Engineers
Engineering

Engineering Team Structure: Models That Scale from 10 to 100 Engineers

Three engineering team structures fit three different company sizes. Pick the wrong one and the next 20 hires arrive into a system that cannot absorb them.

June 9, 2026 · 6 min read
Engineering Velocity: How to Measure It and Why It Stalls
Engineering

Engineering Velocity: How to Measure It and Why It Stalls

Engineering velocity is not the sum of story points. It is the time from decision to production. Here is how to measure it properly and what to do when it stalls.

June 2, 2026 · 6 min read
How to Prevent AI Data Leaks: A Practical Guide for Engineering Teams
Security

How to Prevent AI Data Leaks: A Practical Guide for Engineering Teams

AI assistants can leak your data through prompt injection. Six rules to prevent AI data leaks, what each one stops, and a checklist for your team.

May 30, 2026 · 8 min read
Do AI Coding Tools Actually Make Developers Faster? The 2026 Evidence
AI

Do AI Coding Tools Actually Make Developers Faster? The 2026 Evidence

The only randomized trial on AI coding tools measured a 19% slowdown — while developers believed they were 20% faster. Here is what the evidence shows.

May 27, 2026 · 10 min read
SaaS Product Roadmap: Examples, Templates and the Process We Use
Product

SaaS Product Roadmap: Examples, Templates and the Process We Use

Most SaaS product roadmaps read like project plans, get one missed deadline, and stop being trusted. Here is the three-layer roadmap structure, template and review cadence we use with our clients.

May 26, 2026 · 6 min read
CTO Consulting Services & Advisory: A Complete Guide for Founders and Boards
Leadership

CTO Consulting Services & Advisory: A Complete Guide for Founders and Boards

CTO consulting and CTO advisory are not the same product, even when the same website sells both. Here is what each one delivers, what it costs, and how to choose between providers without losing a quarter.

May 19, 2026 · 7 min read
Part-Time CTO vs Fractional vs Full-Time: Choosing the Right Engagement Model
Leadership

Part-Time CTO vs Fractional vs Full-Time: Choosing the Right Engagement Model

A part time CTO, a fractional CTO and a full-time hire are not interchangeable. Here is what each one means in 2026, where each one earns its price, and how to pick without losing six months.

May 12, 2026 · 6 min read
Fractional CTO Cost: Hourly Rates, Pricing Models and What You Actually Pay For
Pricing

Fractional CTO Cost: Hourly Rates, Pricing Models and What You Actually Pay For

What a fractional CTO actually costs in 2026: the hourly rate ranges, the three pricing models you will be offered, and what drives the total fractional CTO cost up or down.

May 5, 2026 · 6 min read

WEEKLY NOTE

One insight per week. From current engagements.

One short note from current work plus 2–3 outside links worth your time.

Stay updated

Subscribe via RSS or follow us on social media