AI Workforce and Data Science Job Market Trends in the United States

Data & analytics and AI talent are now a binding constraint for enterprise transformation, with GenAI driving demand faster than supply.

This data book maps where U.S. data and AI talent sits, how fast it is growing, what it costs and how the work itself is changing.

Source Draup BrainDesk research·850M+ profiles and 550M+ job descriptions·Focus United States

Data & analytics talent

~0K

employed in the U.S., with ~96,250 job postings and a 5.3% CAGR

AI talent

~0K

employed, with ~60,250 job postings clustered in the same mature hubs

Top hub

0K

data & analytics professionals in San Francisco, the largest U.S. market

Introduction

Data and AI talent is now a binding constraint

Data & analytics and AI talent are now a binding constraint for enterprise transformation, especially as GenAI is driving demand faster than supply. Across the U.S., we see hiring pressure concentrating in a handful of mature hubs, while a second tier of markets offers relief through lower hiring difficulty.

In our latest BrainDesk report, we focused on where data and AI workforces are located, how fast they are growing, what skill mixes are emerging, and what that means for enterprise workforce planning. We also examined how compensation and workflow expectations are shifting as cloud, big data ecosystems, and modern MLOps become baseline requirements. The through-line is clear: location strategy, capability building, and governance maturity must move in lockstep to keep data programs scalable and compliant.

Key data

01

The U.S. employed data & analytics talent pool is ~503,700 with ~96,250 job postings and a 5.3% CAGR (2022–2025), alongside high hiring difficulty.

02

AI talent is also large and clustered: ~317,100 employed with ~60,250 postings. Top hubs include San Francisco (45,700) and New York (39,300), reinforcing the overlap between data platform maturity and AI execution capacity.

U.S. supply–demand reality

Scale is high, but competition is higher

The United States has approximately 503,700 data and analytics professionals, most of whom are concentrated in key metropolitan tech and financial hubs. Even with substantial annual demand (~96,250 postings) and a 5.3% CAGR, the market is still flagged as high for hiring difficulty.

The industry mix shows Enterprise Software (21.7%) and Banking & Financial Services (13.2%) as major anchors, and the experience distribution is weighted toward mid- to late career, which typically raises costs and extends time-to-fill for senior roles.

Data & analytics talent landscape in the United States

Talent attributes

Employed talent~503,700Job postings~96,250May 2024 – Apr 2025Talent CAGR5.3%2022–2025Hiring difficultyHigh

Industry split

Enterprise Software21.7%Banking & Financial Services13.2%IT Services8.4%Professional Services8.1%Consumer Software5.5%Retail4.9%Media & Entertainment4.9%Telecommunications2.7%Product Engineering Services2.5%Insurance2.1%Others*25.9%

Gender & ethnic diversity

GenderWomen 32%Men 68%32%womenEthnic diversity5%Black4%Hispanic35%Asian

Years of experience split

20%0–5 years32%6–10 years21%11–15 years27%15+ years

Talent numbers are extracted from Draup’s database of 850M+ profiles. The employer list is not exhaustive; prominent employers and diversity metrics come from Draup’s proprietary Talent Module. Job postings cover May 2024 – April 2025; growth rate is CAGR for 2022–2025 (Q1). *Other industries include Consumer Electronics, Semiconductors, Logistics & Supply Chain and more.

Source: Draup’s Proprietary Talent Module.

Insights

01

McKinsey predicts that by 2030, generative AI and automation will transform the talent life cycle, shifting roles from basic analytical tasks to higher-level functions such as data product development and lineage production.

02

LinkedIn’s 2025 Job Market Outlook forecasts that data science roles will increasingly demand expertise in model interpretability, responsible AI and bias mitigation to ensure ethical and explainable systems.

03

Deloitte forecasts that U.S. tech jobs will grow from 6 million in 2023 to 7.1 million by 2034, intensifying competition for data engineering and analytics talent.

AI workforce distribution

Top talent hubs and emerging hiring markets

San Francisco and New York lead in data and analytics talent, with mid-tier and emerging markets also hosting notable talent concentrations. The map shows a clear top tier led by San Francisco (63,100) and New York City (51,000), with Seattle (30,100) and Dallas–Fort Worth (28,300) forming a strong second band.

The long tail matters because these markets can support satellite teams or specialized pods if the operating model is designed for distributed delivery. For enterprise HR, we interpret this as a two-speed location strategy: maintain presence in mature hubs for senior leadership and hard-to-source specialties, while building scalable delivery capacity in lower-cost, lower-friction metros.

Data & analytics talent hotspots

San FranciscoNew YorkSeattleDallas-Fort WorthWashington D.C.High favourabilityModerateLowBubble size = employed talent

Bubble size shows employed talent; color shows the favourability score. MSAs include HQ, regional offices and employees’ remote work locations. Hover a city for details.

Source: Draup’s ML model, which analyzes 2M+ publications, industry reports and news articles weekly.

Hiring difficulties

Mature hubs vs emerging options

The dashboard below makes the trade-off explicit. Mature hubs combine high employed talent and postings with consistently high hiring difficulty. In contrast, Austin, Philadelphia, and Denver show low hiring difficulty and can serve as actionable levers for workforce planners.

We also note that growth isn’t exclusive to top hubs; several mid-tier metros sustain ~4–5% growth, which can compound quickly when paired with a strong internal mobility and learning system.

Data & analytics talent dashboard: San Francisco and New York lead, while Austin, Philadelphia and Denver offer lower hiring difficulty

LocationEmployedtalent¹Employed(BFSI)Jobpostings²Growth rate³HiringdifficultyGenderdiversityYears of experience 0–5 |6–10 | 11–15 | 15+United States503,70090,40096,2505.3%32.2%HighSan Francisco63,10011,55010,1006.5%34.2%HighNew York City51,00012,2008,0606.2%33.1%HighSeattle30,1004,8005,3205.1%31.5%ModerateDallas-Fort Worth28,3005,6504,3305%35.5%HighWashington D.C.26,0004,2003,8004.4%30.9%HighLos Angeles19,0003,7503,3205.1%31.1%ModerateChicago18,6003,2002,6903.7%31%ModerateAtlanta15,2002,2502,2505.3%32.4%ModerateBoston14,3001,8002,4004.8%32.9%ModerateAustin10,9001,2001,6004.9%30.1%LowPhiladelphia10,3001,5001,5504.2%31.6%LowDenver9,4001,1001,4004.6%29.4%Low

Cell shading shows the relative level (brighter = higher). ¹Employed talent ²Job postings, May 2024 – April 2025 ³Growth rate, CAGR 2022–2025. Only IT workload-related job roles are considered, filtered on AI-related skills.

Source: Draup’s Proprietary Talent Module.

Compensation signals

Pay premiums in top hubs shape build-vs-buy decisions

The pay gradient is steep and persistent across experience bands. Major hubs like San Francisco, New York, and Seattle lead the way in compensation for Data Scientist roles across experience levels.

The dashboard below supports a portfolio approach: reserve premium-hub compensation for high-leverage roles (platform architects, principal ML, governance leaders) while scaling execution capacity in more favorable markets.

Data Scientist median base pay by MSA and experience

San Francisco$139.1K0–56–1011–1515+New York$135.1K0–56–1011–1515+Seattle$132.4K0–56–1011–1515+Los Angeles$121.8K0–56–1011–1515+Boston$119.8K0–56–1011–1515+Washington D.C.$119K0–56–1011–1515+Philadelphia$113.5K0–56–1011–1515+Chicago$113.4K0–56–1011–1515+Dallas-Fort Worth$111.2K0–56–1011–1515+Atlanta$110K0–56–1011–1515+Detroit$109.8K0–56–1011–1515+Austin$108.8K0–56–1011–1515+0–5 yrs6–10 yrs11–15 yrs15+ yrs95th percentileFigure in panel = median base pay, all bands

All salaries are base salaries and exclude additional compensation and benefits. The 95th percentile is analyzed across the highest-paying firms. Hover a panel for exact values.

Source: Draup’s Cost Simulation Module, using salary platforms, company job postings and official boards, normalized to the Data Scientist role.

AI job market dynamics

Base expectations today

Now that we understand the market, let’s look deeper into job taxonomies and workloads for data and analytics roles. Draup conducted a comprehensive analysis of 850M+ profiles and 550M+ job descriptions to categorize the job role taxonomy for various data & analytics roles.

Data & analytics job roles taxonomy

Data ScienceChief Data ScientistDirector, Data SciencePrincipal ScientistData Science Lead /ManagerSenior Data ScientistStatistical Data ResearchScientistData ScientistStatisticianData EngineeringDirector, Data EngineeringManager, Data Architecture& StrategyData Engineering ManagerData ManagerTechnical ArchitectETL EngineerBig Data EngineerData Modelling EngineerData Integration EngineerData ModelerData Engineer / ArchitectData GovernanceHead, Data OperationsChief Data OfficerData Strategy ManagerData Protection OfficerData Policy SpecialistData Governance AnalystData Security EngineerData Privacy SpecialistData Quality AnalystData StewardMachine Learning andAIDirector, Machine Learning& AIPrincipal Machine LearningScientistSenior Engineering Manager– MLManager, ML InfrastructureEngineeringDeep Learning EngineerComputer Vision EngineerNLP EngineerAI Scientist / EngineerAlgorithm EngineerML Scientist / ResearcherMachine Learning EngineerData AnalyticsDirector, Data IntegrationBI DirectorBusiness IntelligenceManagerManager, AdvancedAnalyticsManager, Data AnalyticsData System AnalystBI AnalystData & Analytics EngineerBusiness AnalystData AnalystComparative role senioritySeniorManagementIC

Roles listed are indicative, not exhaustive, and are shaded by comparative seniority.

Source: Draup’s Proprietary Talent Module (850M+ talent profiles, 550M+ job descriptions).

Data engineering and data science roles are adopting emerging technologies such as cloud platforms, big data ecosystems, and advanced machine learning to keep pace with the evolving digital landscape. Data governance roles are becoming increasingly critical with AI advances, driven by a growing emphasis on privacy and data security compliance.

Data & analytics role skills and workloads

Job family
Traditional functional skills
Emerging functional skills
Key behavioral skills
Workloads
Data Engineering
• Experience in SQL and database architecture
• Analytics development
• Data integration and database management
• Building data pipelines; implementing, testing and maintaining infrastructural components
• Data collection & management
• Presenting data in graphs, charts, tables, etc.
• Troubleshooting complex technical problems on large datasets
• Cloud-based technologies, relational databases, data warehouse, big data (Hadoop, Spark, Hive, Pig)
• Azure analytics stack (Data Lake, Databricks, Data Factory, Synapse, Logic Apps)
• ETL process management
• Building solutions on elastic architectures (Microsoft Azure, Google Cloud Platform)
• A/B testing and analysis of machine learning models
• Strategic decision making
• Reporting
• Communication
• Continuous learning mindset
• Strategic planning
• Organizing
• Team player
• Handling workloads
• Critical thinking
• Problem solving
• Outcome oriented
• Design, develop and maintain data pipelines and back-end and front-end services for reporting, monitoring and analysis
• Ship high-quality, well-tested, secure and maintainable code; continuously improve the tech stack
• Build data processing tools, libraries and frameworks and get insights from data analysis
• Manage CI/CD pipelines and automation for data science projects
• Develop solutions for data quality validation and continuous improvement
• Perform root cause analysis of system and data issues
Data Science
• Programming: Python, R, Java and SQL
• Quantitative modelling, analysis and forecasting
• SQL scripts for analysis and reporting (Redshift, SQL Server, MySQL)
• Visualization with Tableau, Power BI, matplotlib/seaborn
• Data mining and modelling
• Predictive analytics (forecasting, time series, neural networks)
• Machine learning techniques such as GBM, random forest, XGBoost
• Machine and deep learning toolkits: MXNet, TensorFlow, Caffe, PyTorch
• Big data platforms like Apache Spark and Hadoop
• Data flow, pipeline and workflow tools such as Airflow
• Data presentation
• Communication
• Business problem solving
• Business decision making
• Analytical skills
• Presentation
• Efficiency
• Listening
• Interpersonal skills
• Strong interaction
• Effective planning
• Analyze structured and unstructured data to extract insights for model development and customer experience
• Design structured, multi-source data and modelling solutions
• Establish scalable, automated processes for large-scale analyses, model development, validation and implementation
• Run A/B experiments and deliver causal-inference recommendations for product decisions
Data Governance
• Data quality management (profiling, cleansing, monitoring)
• Metadata management & data modelling
• Master data management (MDM) principles
• Policy & standards development
• Data classification & cataloging
• Regulatory compliance reporting (e.g., SOX, HIPAA)
• Risk & controls assessment
• Enterprise data catalog & lineage tools (Collibra, Alation, Informatica EDC)
• Privacy compliance (e.g., CCPA) frameworks & tooling
• AI/ML-driven data quality & anomaly detection
• Cloud-native governance in Azure/AWS/GCP (Purview, Lake Formation)
• Blockchain-based audit trails
• Automation/orchestration (e.g., stewardship workflows)
• Attention to detail
• Ethical judgment & integrity
• Stakeholder management & influence
• Strategic thinking
• Collaboration & facilitation
• Clear communication
• Change-management aptitude
• Analytical problem-solving
• Define, document and enforce data governance policies, standards and procedures
• Maintain the data catalog, metadata repository and lineage maps
• Run data quality assessments and coordinate remediation
• Audit for regulatory compliance and prepare reports
• Train and certify data stewards and business users
• Configure, monitor and optimize governance workflows
• Track KPI dashboards (compliance score, data quality trends)

Emerging functional skills are highlighted. Scroll sideways on smaller screens.

Source: Draup’s database of 1M+ digital intentions from news, job descriptions, video interviews and journals.

Work design is changing

Modern data science becomes end-to-end product delivery

The modern data science workflow has significantly evolved, expanding beyond traditional data processing to incorporate concepts and practices such as AI, ML, and Internet of Things.

The workflow comparison below shows a shift from a linear, analyst-style loop (sourcing → processing → modeling → deployment → monitoring) to an end-to-end lifecycle with explicit capabilities in data acquisition, preprocessing/versioning, model training, versioning/packaging/deployment, AutoML/hyperparameter tuning, and real-time/batch evaluation with monitoring and feedback.

For enterprise HR, this reframes job architecture: high-performing teams need MLOps, platform engineering, and responsible deployment skills, not only model builders. This also strengthens the case for structured internal pathways from data engineering and analytics into ML engineering roles.

Traditional data scientist workflow

Data SourcingRaw source data can be a group ofCSV files or a table in a databaseserverData ProcessingClean the data to make it morestructured and meaningful to themodelData ModellingUse machine learning algorithms totrain the processed data and buildthe modelDeploymentDeploy the model so that it can bequeried for future data so thatpredictions can be madeMonitoringStore the inputs to the model andthe output for creating reports andvisualizing performance

New-age data scientist workflow

Data AcquisitionApache Kafka, AWS Kinesis, Cloudera Morphlines, Databus,Apache Samza, Apache Flume, AWS S3, GlueCollecting data by various frameworks and formatsData Pre-Processing, Transformationand VersioningPython Pandas, R, Azure ML, AWS Batch, Spark, FlinkRefers to techniques to ‘clean’ data by removingoutliers, replacing missing values, smoothing noisydata, and correcting inconsistent data. Transform,enrich and convert the data to appropriate formssuitable for machine learningReal time inference and BatchEvaluation, Monitoring and FeedbackSageMaker, Azure, Google Cloud, Cortex, Python, Jupyter,Greengrass, A/B Test, MLWatcher, Neptune, TensorBoard,Hyperdash, GuildAIGet inferences on a test dataset as a batch or on realdata. Observe the model performance, log it and feedit to future datasetsAuto ML / Hyperparameter tuningAWS Cloud AutoML, Azure, Google Cloud, H2O, OptunaConditional task to tune the hyperparameters of thefactorization machine to find the best modelModel Versioning, Packaging and DeploymentDVC, Docker, Pachyderm, Kubernetes, Kubeflow, AWS,Azure, Google Cloud, IoT, PMML, PFA, ONNXExporting the final ML model and deploying it behind aserver or on IoT devicesML Algo, Deep Learning, Model TrainingAzure, SageMaker, Google Cloud, Python, TensorFlow, Accord.NET, Apache Mahout,Apache Singa, Apple Core ML, C++, R, MXNet, PyTorch, MLlib, AlgorithmsAlgorithms and libraries to build ML models. Process of applying the machine learningalgorithm on training data to train an ML model

The evolution analysis provides a cross-industry view and is not limited to any specific industry.

AI talent at scale

Clustered hubs and fast market expansion raise the urgency

The United States is home to approximately 317,100 AI professionals, primarily concentrated in established hubs, with strong growth in the sector, driven by increasing demand across industries. AI talent is meaningfully smaller than the broader data & analytics pool, and it clusters strongly in the same mature hubs.

With ~60,250 job postings, enterprises that delay capability building will face escalating wage pressure and longer requisition cycles, especially if they require on-site hiring in top hubs.

AI talent in the United States

Talent attributes

Employed AI talent~317,100AI job postings~60,250Gender diversity30.5%women

Top MSAs for AI talent

San FranciscoNew YorkSeattleWashington D.C.BostonHigh favourabilityModerateLowBubble size = employed talent

Years of experience split

24%0–5 years27%5–10 years49%10+ years

Talent numbers are extracted from Draup’s database of 850M+ profiles. Diversity metrics come from Draup’s proprietary Talent Module. Job postings cover May 2024 – April 2025. Bubble size shows employed AI talent; color shows the favourability score.

Source: Draup’s Proprietary Talent Module.

Insights

01

The U.S. AI market reached approximately $146.09 billion in 2024 and is projected to grow to $173.56 billion by 2025, a CAGR of 19.33%.

02

Stargate LLC, a joint venture of OpenAI, SoftBank, Oracle and MGX, plans to invest up to $500 billion in U.S. AI infrastructure by 2029.

03

The U.S. generative AI market was valued at approximately $4.06 billion in 2023 and is projected to grow at a 36.3% CAGR, reaching $33.78 billion by 2030.

Conclusion

From hiring for projects to building durable capability

Across the U.S., we see data, analytics, and AI talent demand scaling in a market where the hardest constraint is not headcount; it’s competition in mature hubs and the evolving skill expectations of production-grade delivery. The most resilient enterprise talent strategies will have to combine a two-tier location footprint, compensation discipline aligned to role criticality, deliberate reskilling pathways into modern ML/MLOps workflows, and governance capacity that scales alongside AI adoption.

In practice, this means shifting from hiring for projects to building durable capability. Designing job architectures for end-to-end lifecycle work, investing in internal mobility, and treating data governance as a first-order operating requirement rather than a compliance afterthought.

Where should your next data and AI team sit?

Compare talent depth, hiring difficulty and pay across U.S. hubs before you commit headcount.