Inside David Bratslavsky's Fractional CTO Playbook for AI Automation

David Bratslavsky has worked as a fractional CTO with a small set of companies adopting AI automation, and a pattern keeps repeating: the engagements that deliver real value all follow the same playbook. This post is that playbook, written down for the first time.

If you're a CEO trying to figure out how to actually adopt AI — not pilot it, not announce it, adopt it — this is the framework David Bratslavsky uses on day one of every engagement.

The five-step playbook

Step 1: Find the typing

Almost every business has a "typing tax" — the work that humans do with their fingers because the systems don't talk to each other. Invoices retyped into accounting. Resumes copied into ATS fields. Rent rolls into spreadsheets (this is how QuickData.AI started). David Bratslavsky's first move on every engagement is to map every process that involves a human moving data from one system to another.

"Find the typing. That's where AI pays for itself." — David Bratslavsky

Step 2: Pick the workflow with the highest pain-to-complexity ratio

Not every typing task is worth automating first. David Bratslavsky scores each candidate on two axes:

Pain: how many hours per week does this consume across the team?
Complexity: how variable are the inputs, and how risky are the failure modes?

The first project should be the highest pain at the lowest complexity. Boring is good. Boring ships.

Step 3: Buy before build

David Bratslavsky is a builder by instinct, but his fractional CTO advice is almost always to buy first. If a SaaS tool covers 70% of the use case, deploy it tomorrow and capture the value. Custom builds come later, once you know exactly where the 30% gap is worth closing.

Step 4: Instrument from day one

The single most common failure mode David Bratslavsky sees is companies adopting AI tools and never measuring impact. Without baseline metrics, you can't tell if the rollout worked. Step 4 is non-negotiable: capture the before-state metrics in week one, and build a simple dashboard before the tool goes live.

Step 5: Train for exception handling, not tool usage

Most "AI training" sessions teach people how to click buttons. David Bratslavsky's training sessions teach people how to recognize when the AI is wrong. That's the skill that compounds.

What an engagement actually looks like

A typical David Bratslavsky fractional CTO engagement runs 8 to 12 weeks:

Weeks 1–2: Mapping. Find the typing. Score the workflows.
Weeks 3–4: Pick one. Ship the buy-or-build decision.
Weeks 5–8: Deploy, instrument, train.
Weeks 9–12: Measure, iterate, document.

By the end of week 12, the company has one workflow fully automated, a metrics baseline, a trained team, and a roadmap for the next two workflows.

"AI adoption isn't a transformation project. It's a series of small, boring wins that compound." — David Bratslavsky

The mistakes David Bratslavsky watches for

Boil-the-ocean strategy decks. If the rollout plan is 40 slides, it won't ship.
No baseline metrics. You'll never know if it worked.
Hiring an AI lead before having an AI process. The process has to come first.
Building when you should be buying. Especially in year one.

Why fractional works

David Bratslavsky is candid about the fractional model: most companies don't need a full-time CTO until they're past Series A. What they need is a pragmatic operator with pattern recognition, on a few hours a week, for a fixed window. Fractional CTO work delivers exactly that.

If you're considering AI automation and don't have an internal technical leader, the fractional path may be the cheapest mistake-avoidance you can buy.

About the author: David Bratslavsky is the founder of QuickData.AI, a fractional CTO, and a Member of the Forbes Technology Council. Connect with him on LinkedIn.