Field Notes

AI Didn't Take My Job.
It Made Me Three People.

Helee Abutbul · April 2026 · 5 min read

If you're a data person, you've probably heard the fear: AI is coming for your job.

I want to show you what actually happened to mine.

In the past month, working with two different clients, I saved $5,000 a month in tool costs — and delivered more impact than I would have with a full team a few years ago. Not because I worked harder. Because AI let me wear hats I never could before.

$2,000 saved / month
pipeline costs
$3,000 saved / month
BI tool costs
−15% churn
improvement

Here's what that looked like in practice.

Hat #1: Data Engineer

The problem

Replacing a $2,000/month Fivetran pipeline

One client was paying Fivetran to pull data from an SFTP server into Snowflake. It worked fine. It was also $2,000 a month for what was, at its core, a straightforward data movement task. I replaced it with a custom pipeline — built, tested, and deployed it myself in days.

$2,000/month saved

A few years ago that would have meant hiring a data engineer or spending weeks on something outside my core skill set. The gap between data analyst and data engineer is closing fast. Custom pipelines, ETL processes, Snowflake integrations — these used to require specialist skills and expensive tools. Increasingly, they don't.

Hat #2: BI Developer + AI Prompt Engineer

The problem

Replacing $3,000/month in Tableau costs

Another client was paying $3,000 a month for Tableau. Good dashboards. But dashboards that told them what happened — not what to do about it. I built them a custom data app combining traditional BI with an LLM layer that interprets the data in plain language. Like having a senior analyst available 24/7, reading the numbers and telling the team what they mean.

$3,000/month saved

BI tells you what the numbers are. AI tells you what they mean. Together, they close the gap between data and decision.

This is what I mean when I talk about AI + BI as a combined discipline. The app costs a fraction of what Tableau cost — and it does more.

Hat #3: Analytics Engineer

The problem

Acting on churn signals before it's too late

Most data teams operate on a lag — by the time you see a churn signal in a weekly report, the customer is already gone. I set up automated insights delivered directly to Slack. When a pattern changes — usage drops, engagement falls, a key signal shifts — the CS team gets a notification in real time. Not a dashboard to check. A message that says: this customer needs attention, now.

Churn improved by 15%

This is the difference between reporting and intelligence. Reporting tells you what happened. Intelligence tells you what to do, when it still matters.

What this actually means for data people

I'm not saying pipelines don't matter or that BI tools are dead. They're not. Invest in infrastructure — but measure yourself by the decisions you empower, not the pipelines you ship.

What I am saying is that the definition of what one data person can do has fundamentally changed. A few years ago, the data stack required specialists — a data engineer for the pipelines, a BI developer for the dashboards, an analyst for the insights. Each with their own tools, their own costs, their own timelines.

Today, a data consultant with the right skills and the right AI tools can cover all of that. Not perfectly. Not at enterprise scale. But well enough to save a growing company $5,000 a month and actually move business metrics.

Data people aren't at risk from the AI revolution — they're at the center of it. The ones who see that will be the ones leading their organizations through it.

Is your data stack costing more than it should?

Let's look at where your investment is going — and whether it's actually driving decisions.

Let's talk →