AI/ML Job Market Analysis — Siddharth Sadhu

Live Preview

The Dashboard

A single-page interactive Power BI dashboard with 7 visuals, 4 slicers, and 9 DAX measures — built for both class presentation and portfolio showcasing.

app.powerbi.com · AI / ML Job Market Analysis — Dashboard · Siddharth Sadhu · Adani University

AU

ADANI UNIVERSITY

AI / ML Job Market Analysis — Dashboard

⬡ LIVE DEMO ↗

147.00K

Median Salary

800.00K

Max Salary

151K

Total Jobs

20.76

Remote Jobs %

Top 10 Highest Paying AI/ML Roles

Research Team Lead

Analytics Eng...

Data Science Lead

Applied AI/ML Lead

Head of ML

IT Enterprise DA...

Salary Growth Trend by Experience (2020–2025)

Executive Senior Mid Entry

Salary Band Distribution

Avg Salary by Experience

Exec

Senior

Mid

Entry

Global AI/ML Job Distribution

Avg Salary by Company Size

Large$158K

Medium$142K

Small$98K

Work Year: 2020–2025

Experience: All Levels

Company Size: All

Work Mode: All

Data Source

Dataset Details

Complete information about the dataset — its origin, structure, size, and every column explained.

CSV

Data Science, AI & ML Job Salaries 2025

kaggle.com/datasets/adilshamim8/salaries-for-data-science-jobs ↗

Total Records151,000+ rows

Columns Used9 (cleaned from 11)

Year Range2020 – 2025

LicenseCC0 Public Domain

FormatsCSV JSON

Data SourcesAIJobs, Payscale, KDnuggets, ZipRecruiter

Geographic ScopeGlobal · 60+ countries

CurrencyAll normalized to USD

Columns Removedsalary (raw) salary_currency

Real-world dataMulti-country 2020–2025AI/ML focused 151K+ recordsFree on Kaggle

Download Dataset ↗ View Live Dashboard ↗

9 Column Reference

Work Year

Whole Number

Year the salary was recorded. Ranges from 2020–2025. Used as X-axis in the salary trend line chart.

Experience Level

Text · Entry/Mid/Senior/Executive

Career stage. Decoded from EN/MI/SE/EX in Power Query. Primary slicer and analysis dimension.

Employment Type

Text · Full-Time only

Filtered to Full-Time for consistent benchmarks. PT/CT/FL records removed in Power Query.

Job Title

Text · 80+ unique roles

Specific role name. ML Engineer, Data Scientist, AI Researcher etc. Used in Top 10 bar chart.

Salary (USD)

Decimal · USD normalized

All salaries normalized to USD. Raw multi-currency column removed to prevent incorrect aggregation.

Employee Country

Text · ISO code

Country where the employee lives. May differ from Company Country for remote workers.

Work Mode

Text · Onsite/Hybrid/Remote

Decoded from 0/50/100. Powers the Remote Jobs % KPI card measure.

Company Country

Text · ISO code

Where the hiring company is based. Used in map visual for geographic job distribution.

Company Size

Text · Small/Medium/Large

Decoded from S/M/L in Power Query. Used in salary comparison bar chart and slicer.

Findings

Key Insights

Six data-backed discoveries from 151K+ records that decode the AI/ML job market.

01 · TOP ROLES

Research Team Lead commands the highest compensation

Leadership AI roles significantly outpace pure engineering positions, signaling that cross-functional expertise is valued above deep technical skill alone at the senior level.

$350K+

02 · EXPERIENCE

Executive earns nearly 2× entry-level salary

Experience is the strongest single predictor of compensation in AI/ML — outweighing geography, company size, or specific job title in most cases across all regions.

~2× gap

03 · REMOTE WORK

Only 20.76% of AI/ML jobs are truly remote

Despite the remote-first tech narrative, the majority of AI/ML roles remain onsite or hybrid — suggesting lab access, hardware requirements, or team collaboration needs.

20.76%

04 · SALARY BANDS

67% of all roles fall in the $60K–$120K mid band

The workforce concentrates in the mid range. The High band (28.36%) is reserved primarily for Senior/Executive roles at large companies, especially in North America.

67% mid

05 · COMPANY SIZE

Large companies pay 61% more than small firms

Large orgs average $158K vs $98K at small companies. However, some startups match or exceed this for niche AI specializations like LLM research or autonomous systems.

+61%

06 · GROWTH TREND

Salaries have grown every year from 2020–2025

The line chart shows consistent growth across all experience levels, with the sharpest acceleration occurring post-2022 — coinciding directly with the generative AI hiring surge.

↑ 5yr trend

Dashboard Guide

How to Interpret the Dashboard

A visual-by-visual guide — what each chart shows, how to read it correctly, and how slicers interact.

KPI

The 4 KPI Cards (Top Row)

Show Median Salary, Max Salary, Total Jobs, and Remote Jobs % for the currently filtered data. All four update dynamically with every slicer click. Median is used (not average) to avoid distortion from extreme executive salaries pulling the number unrealistically high.

Try: Click "Entry Level" in the Experience slicer to benchmark what a fresh AI/ML graduate earns.

BAR

Top 10 Paying Roles (Bar Chart)

Ranked by Average Salary (USD) with a Top N filter showing only the top 10 titles. Longer bar = higher average. Job titles may be truncated — hover to see the full name. Changing slicers will shift which roles appear and their rankings.

Try: Filter to "Large" company + "Senior Level" to see elite compensation benchmarks.

LINE

Salary Growth Trend (Line Chart)

Shows average salary over time (2020–2025) broken by experience level. Four lines = Entry, Mid, Senior, Executive. The widening gap between lines over time shows senior professionals have benefited most from AI demand growth.

The sharp uptick from 2022–2023 correlates directly with the ChatGPT-driven AI hiring surge.

Salary Band Distribution (Donut Chart)

Categorizes all records into Low (<$60K), Mid ($60K–$120K), and High (>$120K) using the DAX calculated column "Salary Band". The 67% Mid figure shifts when filtering to Senior+ only. This is NOT a raw data field — it was created with nested IF statements in DAX.

This chart demonstrates the Salary Band calculated column — mention this in your viva.

MAP

Global Job Distribution (Map)

Bubble size = number of job records per country. North America and Europe dominate. Note: shows Company Country (employer location), not Employee Country, which may differ for remote workers. US concentration is why the median salary ($147K) appears high globally.

Select "Remote" in Work Mode slicer to see how geographic distribution changes for remote roles.

SLI

4 Slicers — Interactive Filters

Work Year, Experience Level, Company Size, and Work Mode all control every visual simultaneously. Cross-filtering is fully enabled — clicking a bar in any chart also filters all other visuals. Use combinations like "Remote + Entry + Large" for targeted analysis.

Hold Ctrl and click multiple options in a slicer to select more than one value at a time.

Visual Inventory

All 9 Visuals & Components

Every visual used — its type, data fields mapped, and the story it tells.

Card Visual × 4

KPI Cards

Median Salary, Max Salary, Total Jobs, Remote Jobs %. DAX-powered. Update with every slicer interaction.

Clustered Bar Chart

Top 10 Paying Roles

Y: Job Title · X: Avg Salary · Top N filter = 10. Benchmark for role-level compensation.

Line Chart

Salary Trend by Experience

X: Work Year · Y: Avg Salary · Legend: Experience Level. Shows 5-year career salary growth.

Donut Chart

Salary Band Distribution

Legend: Salary Band (calculated column) · Values: Total Jobs. Low / Mid / High tier breakdown.

Clustered Column Chart

Avg Salary by Experience

X: Experience Level · Y: Avg Salary. Clear visual proof experience = #1 salary driver.

Map Visual

Global Job Distribution

Location: Company Country · Size: Total Jobs. Geographic concentration of AI/ML employment.

Clustered Bar Chart

Avg Salary by Company Size

Y: Company Size · X: Avg Salary. Compares Small, Medium, Large org compensation.

Slicer × 4

Interactive Filters

Work Year (range slider), Experience Level, Company Size, Work Mode. All enable cross-filtering.

Text Box + Image

Header & Branding

Adani University logo, dashboard title, Live Demo button. Professional page anchoring elements.

Process

Project Methodology

Step-by-step walkthrough of how this was built — from raw CSV to polished, interactive dashboard.

Data Acquisition

Dataset Selection & Import

Selected the Kaggle AI/ML Salaries 2025 dataset for real-world relevance, recency, and domain alignment with the AI/ML career trajectory.

Identified CC0 licensed dataset with 151K+ records

Downloaded CSV from Kaggle

Loaded via Power BI's native Get Data → CSV connector

Power Query

Data Cleaning & Transformation

Used Power Query Editor to clean, decode, and reshape the raw dataset into an analysis-ready format before loading into the model.

Removed raw salary + salary_currency columns

Decoded EN/MI/SE/EX → readable labels

Decoded 0/50/100 → Onsite/Hybrid/Remote

Decoded S/M/L → Small/Medium/Large

Renamed all 9 columns for readability

Filtered to Full-Time employment only

Data Modeling

Model Structure & Measures Table

Single flat table model — all dimensions embedded in the fact table. A dedicated _Measures table separates all DAX calculations from raw data columns.

Single table — no relationships needed

Created _Measures table (professional best practice)

Used Dummy column placeholder to keep table visible

DAX

Measures & Calculated Columns

Built 7 measures and 2 calculated columns covering aggregations, conditional filtering, and salary categorization using CALCULATE, DIVIDE, and nested IF.

7 measures in _Measures table

2 calculated columns on fact table

CALCULATE used for conditional filter overrides

DIVIDE used for safe % calculation

Visualization

Dashboard Layout & Design

Designed a single-page layout with structured visual hierarchy — header, KPI row, two chart rows, and interactive slicer panel. Visual types chosen to best represent each data story.

7 visuals + 4 slicers on single page

Consistent titles, labels, legends across all visuals

Cross-filtering enabled on all visuals

Logo and Live Demo button in header

Review & Polish

Validation & Final Formatting

Reviewed all visuals for accuracy, updated deprecated visual types, verified slicer interactions, formatted charts with proper titles and labels.

Replaced deprecated Card visual with new version

Fixed map visual retirement warning

Tested all 4 slicers for cross-filter behavior

Final layout alignment using Power BI align tools

Tech Stack

Power BI DesktopPower Query EditorDAX Kaggle DatasetCSV Format Azure Maps VisualData Modeling

DAX Reference

All DAX Measures & Columns

Every formula used — with syntax highlighting and explanation of logic and purpose.

Avg SalaryMeasure

Avg Salary = AVERAGE(AI_Job_Salaries[Salary (USD)])

Average salary across all filtered records. Updates dynamically with every slicer interaction.

Median SalaryMeasure · KPI Card

Median Salary = MEDIAN(AI_Job_Salaries[Salary (USD)])

Primary KPI metric. Median preferred over average — less affected by extreme executive outliers that distort the true typical salary.

Max SalaryMeasure

Max Salary = MAX(AI_Job_Salaries[Salary (USD)])

Returns the single highest salary in the currently filtered dataset. Shows the compensation ceiling.

Total JobsMeasure

Total Jobs = COUNTROWS(AI_Job_Salaries)

Counts rows in current filter context. Reflects the sample size powering all other calculations.

Remote Jobs %Measure · Advanced

Remote Jobs % = DIVIDE( CALCULATE( COUNTROWS(AI_Job_Salaries), AI_Job_Salaries[Work Mode] = "Remote" ), COUNTROWS(AI_Job_Salaries), 0 ) * 100

CALCULATE applies an additional filter (Remote only) inside the existing context. DIVIDE handles zero-division safely when no rows match.

Senior+ Avg SalaryMeasure · Advanced

Senior+ Avg Salary = CALCULATE( AVERAGE(AI_Job_Salaries[Salary (USD)]), AI_Job_Salaries[Experience Level] IN {"Senior Level", "Executive Level"} )

Overrides external filter context to isolate Senior and Executive levels — demonstrates advanced CALCULATE usage with IN operator.

Entry Level Avg SalaryMeasure

Entry Level Avg Salary = CALCULATE( AVERAGE(AI_Job_Salaries[Salary (USD)]), AI_Job_Salaries[Experience Level] = "Entry Level" )

Benchmarks fresh graduate earning potential. Compare against Senior+ Avg to quantify the career growth opportunity over time.

Salary BandCalculated Column

Salary Band = IF(AI_Job_Salaries[Salary (USD)] < 60000, "Low (< $60K)", IF(AI_Job_Salaries[Salary (USD)] < 120000, "Mid ($60K–$120K)", "High (> $120K)"))

Computed row-by-row, stored in the table. Enables the Donut Chart's categorical salary tier analysis across the dataset.

Is SeniorCalculated Column

Is Senior = IF( AI_Job_Salaries[Experience Level] IN {"Senior Level", "Executive Level"}, "Senior+", "Junior/Mid" )

Binary flag for senior/non-senior segmentation. Enables quick slicer-based filtering between career tiers.

FAQ

Frequently Asked Questions

Common questions about this dashboard, the data, and the methodology — including likely viva questions.

Why is Median Salary the KPI instead of Average?+

Salary distributions are right-skewed — a small number of executive roles earning $500K–$800K pull the arithmetic mean significantly upward, making it appear higher than what most professionals actually earn. The median is the midpoint of all salaries and represents what a "typical" AI/ML professional earns far more accurately. This is the same methodology used by government salary surveys and HR benchmarking companies worldwide.

Why was the raw salary column removed?+

The raw salary column contains values in 40+ different currencies — USD, INR, EUR, JPY, and more. If used in any aggregation, Power BI treats all numbers as if they share the same unit, producing completely meaningless results. For example, $80,000 USD and ₹80,00,000 INR would be treated as equal. The salary_in_usd column normalizes all values using real exchange rates, enabling accurate cross-country comparison. Keeping the raw column risks someone accidentally using it and generating incorrect charts.

The median shows $147K — is that a realistic global salary?+

Yes — but with important context. The dataset is heavily weighted toward US-based companies, which dominate the global AI/ML hiring market and pay among the world's highest salaries. For India-based roles, figures are significantly lower. This is exactly why the Company Country filter and map visual exist — use them to filter for your specific region to get locally relevant benchmarks. The global median figure is most useful for understanding the overall market ceiling rather than local expectations.

What is the difference between a Measure and a Calculated Column?+

A Calculated Column is evaluated row-by-row at data load time and stored permanently in the table — like adding a new Excel column. It exists in the model and consumes memory. A Measure is computed dynamically at query time based on the current filter context — it recalculates every time you interact with a slicer or visual. Measures are more memory-efficient and are the preferred approach for aggregations like sums, averages, and percentages. In this project, Salary Band and Is Senior are calculated columns; everything else is a measure.

How does CALCULATE work, and why is it needed for Remote Jobs %?+

CALCULATE is the most powerful function in DAX. It evaluates an expression while simultaneously modifying the filter context — adding or overriding filters on top of whatever slicers are already active. In Remote Jobs %, we need to count only rows where Work Mode equals "Remote" regardless of what the Work Mode slicer says. Without CALCULATE, there is no way to apply conditional filters inside a measure. DIVIDE is then used instead of the division operator to gracefully handle cases where the denominator is zero.

How would you extend this into a Machine Learning project?+

This dataset is an ideal foundation for a salary prediction model. Using Experience Level, Job Title, Company Size, Work Mode, and Company Country as input features with Salary (USD) as the target variable, a regression model using Random Forest or XGBoost could predict expected compensation for any role profile. Feature importance analysis would likely confirm what this dashboard already shows — Experience Level is the dominant predictor, followed by Job Title and Company Country. This connects directly to ongoing work in SaralAI and future ML projects.

Siddharth Sadhu

B.Tech ICT · Adani University

Ahmedabad, Gujarat · 2024–2027
CGPA: 8.77

siddharthsadhu28@gmail.com LinkedIn — Siddharth Sadhu ↗ GitHub — siddharthsadhu ↗

About the Author

Pre-final year developer targeting AI/ML & Data Science roles

I'm a Backend-focused Software Developer with a CGPA of 8.77 at Adani University, building full-stack applications and applying machine learning in real-world contexts. This Power BI project is part of my Data Visualization coursework — but more importantly, it's a genuine deep-dive into the industry I'm actively entering.

I chose the AI/ML job market dataset intentionally. Understanding which roles pay the most, how salaries have grown post-2022, and what percentage of jobs are truly remote gives me strategic clarity as I approach the final stretch of my degree and the start of my career.

Notable work includes SaralAI (a multilingual NLP platform), an SSIP-selected Smart Glasses assistive system, and the Deloitte Data Analytics Virtual Experience. My belief: the best projects are ones you'd build even if there was no grade for it.

LinkedIn ↗ GitHub ↗ Email ↗

PythonFastAPIReact.jsPower BI Machine LearningDAXMySQLNode.js SSIP SelectedCGPA 8.77Deloitte Analytics

AI / ML Job MarketAnalysis Dashboard