Title Beats Tenure: Entry-Level ML Engineers Out-Earn Senior Data Analysts (Live Salary Data)
Analyze This! is our series of live QuantumLayers demonstrations. We take a public dataset, load it straight into QuantumLayers, and let the platform chart it and test it. Nothing on this page is drawn by hand or typed in from a spreadsheet. This time we loaded 151,445 AI, machine learning and data salaries and asked a narrower question: is it better to be senior, or to have the right job title?
The data is the open salary dataset from aijobs.net, published by foorilla on GitHub. Each row is one person’s salary for one year, with their experience level, job title, country and work arrangement. It is already a public dataset in QuantumLayers, so every chart on this page is fetched live when you open it, and each one has a button that opens it as a standalone ql-chart.
Rather than survey the whole market, this post follows a single thread through it. Four steps.
STEP 1Two ladders, very different heights
Take the two titles that sit at opposite ends of the pay table and follow each one up the experience levels. A machine learning engineer starts at an average of $150,331 and climbs to $232,955 at the executive level. A data analyst starts at $88,807, peaks at $123,500 as a senior, and the 110 analysts recorded at the executive level average a little less than that, at $115,112.
Put the two ladders side by side and the bottom rung of one is higher than the top rung of the other. The average entry-level ML engineer earns about $27,000 more than the average senior data analyst. Averages can be pulled around by a few large salaries, so we checked the medians too: $139,650 for entry-level ML engineers against $119,555 for senior analysts. Roughly two in three entry-level ML engineers earn more than the typical senior analyst.
Figure 1: Average salary in US dollars by experience level for machine learning engineers (8,887 records) and data analysts (13,779 records), fetched live from QuantumLayers. Source: aijobs.net salary dataset.
▶ ML engineers on QuantumLayers▶ Data analysts on QuantumLayers
STEP 2Where the first job pays
If title matters this much, it should already show up on day one. Filter the dataset to entry-level records only (13,663 of them) and compare the ten most common titles. Data analyst is by far the most common first job, at 38% of all entry-level records, and it is also the lowest paid of the ten. Data engineers, general analysts, BI analysts and data scientists cluster just above it, between about $92,000 and $105,000.
Then the chart opens up. Entry-level software engineers average about $123,000, ML engineers about $150,000, and research scientists about $193,000. That last figure probably says more about who gets hired into research roles than about the roles themselves: many of these “entry-level” scientists likely join with a doctorate already done, so it is their first industry job rather than their first job.
Figure 2: Average entry-level salary in US dollars for the ten most common entry-level titles, fetched live and sorted in your browser. Source: aijobs.net salary dataset.
▶ All entry-level titles on QuantumLayers
STEP 3The new title is the cheaper one
“AI Engineer” barely existed in this dataset until recently. It appears once in 2022, 47 times in 2023, 571 times in 2024 and 1,394 times in 2025. The dataset itself grew over those years, so the fairer measure is share: AI engineers went from about 1 in 180 records in 2023 to about 1 in 56 in 2025. Over the same stretch, machine learning engineers fell from 1 in 8 records to about 1 in 25.
You might expect the newer, more fashionable title to carry a premium. It does the opposite. At the mid level, AI engineers average $164,853 against $178,783 for ML engineers, about 8% less. At the senior level the gap widens to about 12%: $183,055 against $207,637. One likely reason is that the two titles describe different work. “AI Engineer” is often used for building products on top of existing models, through APIs, retrieval and agents, while “ML Engineer” more often means training and serving the models. The data cannot tell us which, but either way the title on the posting is attached to a different price.
Figure 3: Average salary in US dollars for AI engineers and ML engineers at the mid and senior levels, fetched live. Entry and executive levels are left out because there are fewer than 100 AI engineer records at each. Source: aijobs.net salary dataset.
▶ AI engineers on QuantumLayers
STEP 4Stress-testing the result
A finding like “title beats tenure” is easy to overstate, so we tried to break it.
Is it just location? ML engineers are a little more likely to live in the US (91% of records) than data analysts (86%), and US salaries are higher. So we repeated the comparison with US residents only. The gap barely moves: entry-level ML engineers average $158,054 and senior data analysts $127,101.
Is it a few outliers? No. The medians tell the same story as the averages, and the overlap is limited: about 65% of entry-level ML engineers earn more than the median senior data analyst.
Does it hold across the whole market? QuantumLayers’ analysis of variance flagged job title as the strongest factor in the dataset, and across all 151,445 records it explains about 20% of the variation in salary, against about 10% for experience level.
Where should you be careful? Some of the cells are small. There are 110 executive data analysts, 80 entry-level AI engineers and 58 executive AI engineers in the file, which is why we kept the last two out of Figure 3. A difference built on a few dozen salaries can move a lot when next quarter’s data arrives.
What this means for hiring
If a job title carries this much of the price, it also carries a lot of risk. Put the wrong title on a posting and you either overpay or attract the wrong people. Screen candidates by their last title and you miss the analyst who has been shipping models for two years. A few practical consequences:
- Choose the title deliberately. Whether a role is called AI engineer, ML engineer or analyst sets the pay range before anyone reads the description, so decide it from the work the role actually does.
- Look at what people have built. A candidate’s projects, code and portfolio say more about fit than the title on their last offer letter.
- Benchmark by title and level together. A single average for “data roles” hides a gap of about $100,000 between data analysts and research scientists at the entry level alone.
- Agree on pay early. Sharing the salary range in the first conversation keeps a $150,000 engineer and a $120,000 budget from meeting three interviews too late.
The fine print
This dataset leans heavily toward the United States and toward AI, machine learning and data roles. Its publisher normalizes job titles, and began adding many more software and general engineering titles in 2024, so changes in title counts reflect both the market and what the dataset covers. That is why we compare shares rather than raw counts in Step 3. The figures are salaries converted to US dollars by the publisher, so equity and bonuses, which can be large in ML roles, are not part of the comparison.
How this was built
- Import. The aijobs.net CSV was loaded into QuantumLayers from its raw GitHub URL and set to public. QuantumLayers typed all 11 columns and recognized the coded fields, such as experience level, as categories.
- Analyze. Every chart is filtered to a single job title or a single experience level, so each one is computed over every matching record in the file. The analysis of variance across the whole dataset pointed us at job title as the biggest factor, and the follow-up comparisons in Steps 1 to 3 came from there. Medians, shares and the US-only check were verified against the full file.
- Publish. Each chart is drawn with Chart.js from unauthenticated
POSTrequests to QuantumLayers’ql_get_chart_dataendpoint, and Figures 1 and 3 combine two live requests each. The purple buttons open the same queries on QuantumLayers’ hostedql-chartpage.
Want to check your own title? Open any chart above, change filter_category_value to another job title, and see where its ladder sits.
Data: aijobs.net salary dataset, published by foorilla on GitHub. Figures reflect the dataset at the time of writing; the live charts update when the dataset is refreshed.
This post is part of Analyze This!, a series of live QuantumLayers demonstrations. For something completely different, try 32,540 horror movies or 45,716 meteorite landings. To load a dataset of your own and get the same charts and findings, start building at quantumlayers.com.
