St. Petersburg, Florida 22+ yrs IT · 15+ yrs SEO & digital marketing
Experience built in the real world
22+
Years in technology
Helping businesses build, improve and troubleshoot the technology they rely on.
15+
Years in SEO & search
Helping companies get found online, compete in search and turn visibility into growth.
50+
People · team built and led
Built and led an SEO and web team that grew from 3 to 50 people.
120+
Websites delivered per year
Grew from 20 to more than 120 websites a year as sales scaled — while processes, quality and QA got stronger, not weaker.
Problems
Sentences I hear most
Each one is a symptom. The engagement starts by finding out what it is a symptom of.
01Search
“Organic growth is stuck.”
SEO exists, the reports are green, traffic and leads refuse to scale.
02AI
“Nobody owns AI search.”
Everyone discusses AEO and GEO. Nobody owns measurement, process, or the decision to stop.
03People
“The team doesn’t scale.”
More headcount stopped producing more quality output. The constraint moved somewhere else.
04Automation
“Too much work is manual.”
Reporting, QA and onboarding consume the capacity delivery was supposed to get.
05Operations
“Web operations are chaotic.”
SEO, developers and marketing run as three separate systems with three separate truths.
06Leadership
“We need a second opinion.”
Before a big hire, a redesign, a migration or an AI program, someone senior should disagree properly.
Search outcomes
Selected search growth outcomes
Search strategy translated into measurable business growth, audience development and scalable delivery.
$1B+
In sales · from early stage to scale
Led search strategy as a digital business grew from roughly 50 users to more than $1 billion in sales.
Search was one part of that growth, not the whole of it — the work was making organic discovery hold up as the business scaled.
Nearly 100K
Monthly visitors · owned media
An independent publication grown from launch to nearly 100,000 monthly visitors.
~10K
YouTube subscribers
An audience built from zero through useful, discoverable video, feeding a paid education platform.
Nearly 300
Clients · search at portfolio scale
Search processes, quality controls and reporting held consistent as the client base grew to nearly 300.
The method
Diagnose the system, not the symptom.
One chain runs through every engagement. A break anywhere surfaces as a symptom somewhere else — usually somewhere convincing. Pick a symptom:
Where it looked broken: Web. Where it actually broke: Data.
The chain
Search
WebExpected
DataActual
AI
Automation
Reliability
Operations
People
What it looked like
The dashboard is wrong — the pipeline must be broken.
What it actually was
Fresh API data was being compared against finalized stored data.
Nothing was broken. Two sources were answering slightly different questions, and the reconciliation rule had never been written down. The fix was a data contract, not an engineer.
Where it looked broken: Web. Where it actually broke: Reliability.
The chain
Search
WebExpected
Data
AI
Automation
ReliabilityActual
Operations
People
What it looked like
Every site in the portfolio went down at once.
What it actually was
The verification system had lost connectivity — the sites were fine.
A monitor that cannot detect its own failure will confidently report a catastrophe. Abstention states and a fleet-level canary came out of this one.
Where it looked broken: Automation. Where it actually broke: Operations.
The chain
Search
Web
Data
AI
AutomationExpected
Reliability
OperationsActual
People
What it looked like
We need to automate the workflow. All of it.
What it actually was
Most of the task portfolio was not worth automating at all.
3,089 recurring tasks measured by frequency, duration, variance and risk. Most could run unattended once the rules were written down; some needed a human gate; a small share was simply deleted.
Engagements
Four ways in
Most engagements start with a diagnosis and stop when the system holds without me. Not every one goes all the way down the ladder.
Search, site, data, workflow, automation and team examined as one system. Ends in a constraint map and a 30/60/90 plan.
02
Design
Systems Sprint
One high-impact system rebuilt in weeks, not quarters — reporting, onboarding, AI-search measurement, delivery workflow.
10+ hours → minutesWorkflow simplified
03
Lead
Fractional Advisory
A senior operator on the inside: roadmap, manager coaching, hiring, vendor review, escalation support, executive reporting.
04
Build
AI, Data & Automation Implementation
Built, deployed and verified — dashboards, agent workflows with scoped permissions, monitoring, integrations, rollback.
The model
Operator · Product Lead · Builder
Consulting usually fails in the handoffs between these three. The differentiator isn’t any one of them — it’s translating between all three without losing the intent.
Operator
Owns delivery, capacity, escalations and quality — the part that has to survive Monday morning.
Teams · Delivery · SEO · Escalations · Capacity · QA · Hiring
Proof: built and led an SEO and web team from 3 to 50 people.
Product Lead
Turns recurring operational pain into specifications, roadmaps and systems people actually use.
Governed automation, agents that know when to stop, fail-closed defaults — and a person still accountable at the end of the chain.
Automation judgment
Most work can be automated.
Recurring work is where teams lose their best hours: audits, checks, reports, reviews. Almost all of it can be handled by software — once someone decides what is actually worth doing.
The work is deciding what deserves to run at all, then handing as much of it as possible to software — with a human gate only where a wrong answer is expensive.
Automate what earns its place. Delete the rest.
Mostrecurring work — audits, checks, reports, monitoring — can run without a person once the rules are written down.
Someneeds a human gate before anything ships — approvals, judgment calls, anything expensive to get wrong.
A littleshouldn’t be automated at all — because it shouldn’t exist. Deleting it beats optimizing it.
The split is measured per organization before anything is built — never assumed from a benchmark.
Lab
Things I build to solve real problems
Not a technology list. Software I needed, built on my own time, and still use.
Every morning it collects 300+ sources — sites, Reddit, LinkedIn, X — strips the duplicates, clusters what is actually the same story, ranks it, and writes a short briefing I can read with coffee — in a format built for reading, not for a database.
Where the AI actually earns its place
Fetching, deduplicating and ranking are ordinary code — cheap, testable, predictable. The model is used only for the two jobs that need judgment: deciding what a cluster is really about, and saying it in a sentence.
Runs unattended
Fails visibly, never guesses
Read daily since day one
Pipeline
Collect — 300+ feeds: sites, Reddit, LinkedIn, X
Normalize — one shape, duplicates dropped
Cluster — the same story, once
Summarize — the model’s only job
Deliver — a briefing worth reading
Screenshots · click to enlarge
Scheduled · deterministic first
Push-to-talk waveform
VoicePolisher
Speech → transcription → language cleanup, built by a non-native English speaker for his own daily writing.
1,043 dictations · 78 days
Training load over time
Training log
A self-hosted fitness tracker: workouts, sets and body metrics logged in seconds, with progression and training load worked out for me instead of guessed.
Used every session · progression auto-calculated · no subscription
Abstention, fail-closed behavior and rollback as product features — including the pipeline that changed nothing for days, correctly. 30 minutes, or 45 with the Q&A everyone always has.
“Most digital problems don’t start where the symptom appears.”
Twenty years across infrastructure, code, websites, owned media, SEO, content, team leadership, agency operations, data systems and AI agents. The advantage is not doing everything. It is seeing how one layer breaks another.
Still writing code most weeks
— which is why the advice survives contact with the people who have to implement it.
Russia → China → New York → Southeast Asia → Florida
Write it out in a few paragraphs — no call required first. I’ll tell you whether it’s a problem I can help with, whether you need me at all, and which engagement fits.
1You write a few paragraphs about the symptom.
2Within two business days I reply with an honest read.
3If it fits, we agree a scope — most clients start with an audit.
A written brief gets a considered reply, not a sales call Email hello@alexzuev.com for short questions Remote-first · St. Petersburg, Florida Working across US and international teams
