Data Science Career
The Jupyter notebooks, the stakeholder who wants AI on everything, and the 80 percent of the job that's just cleaning data. The real numbers, the skills that matter, and what data scientists say when the model accuracy plateaus at 72 percent.
How Much Do You Actually Make?
The median is $108,000. Tech companies in SF, NYC, and Seattle push senior roles well above $200K with equity. But 'data scientist' has become an inflated title: some roles are really analyst positions at $65,000, while others are ML engineering roles at $250,000+. The title tells you less than the actual work.
Total compensation at tech companies includes base salary, equity (RSUs), and bonus. A $160K base at a FAANG company might be $250K+ total comp. Non-tech industries (healthcare, retail, finance) pay 20-40 percent less but offer more stability. The ML engineering path typically pays more than the analytics-focused path.
What Do You Actually Do All Day?
The expectation: building AI models that transform the business. The reality: 80 percent of the job is finding, cleaning, and understanding data. The other 20 percent is building something and then explaining to a stakeholder why 72 percent accuracy is actually good.
How to Get In
Bachelor's Degree (4 years)
Computer science, statistics, mathematics, or a quantitative field. Some data scientists come from physics, economics, or engineering. The math foundation matters more than the specific major.
Master's or PhD (2-6 years, common but not required)
Most data scientist roles at competitive companies prefer a master's or PhD. Programs in data science, statistics, computer science, or ML are most common. Self-taught paths exist but face more hiring friction.
Build a Portfolio
Kaggle competitions, personal projects, open source contributions, and blog posts demonstrating analytical thinking. The portfolio matters more than the degree for getting interviews.
First Data Role
Data analyst, junior data scientist, or ML engineer. Some enter through analytics and transition to data science. Others enter through software engineering and add ML skills.
Alternative paths: Data analysis is a more accessible entry point (bachelor's degree, SQL, Python, visualization tools) that can lead to data science with additional statistical and ML learning. Bootcamps (Galvanize, Metis, General Assembly) offer 3-6 month immersive paths. Self-taught data scientists exist but face steeper hiring barriers at top companies.
Job Outlook
The BLS projects 36 percent growth through 2032, one of the fastest-growing careers in the economy. Every industry is generating more data than it can analyze.
Growing sectors: ML engineering, AI/LLM applications, healthcare AI, fintech, and privacy-focused data roles are all expanding rapidly. MLOps (putting models into production) is a particularly hot specialty.
Challenges: Basic BI and reporting work (dashboards, SQL queries) is being commoditized by tools like Tableau, Looker, and AI assistants. Data scientists who only do descriptive analytics face downward pressure.
Technology shift: LLMs and generative AI are changing the field rapidly. Data scientists who can fine-tune models, build RAG systems, and evaluate AI outputs are in extreme demand. Those who only know traditional ML are scrambling to upskill.
Honest Pros and Cons
The Good
- 36% growth, massive demand
- High salaries, especially in tech
- Intellectually stimulating work
- Remote work is very common
- Cross-industry applicability
- Working at the cutting edge of AI/ML
The Hard Truth
- 80% of the job is data cleaning
- Stakeholders often don't understand what you do
- Imposter syndrome is pervasive
- PhD culture can create credential pressure
- Models get built and never deployed
- The field changes so fast that skills decay quickly
Career Paths
Data Analyst
The entry ramp. SQL, dashboards, business questions. Less modeling, more communication.
Data Scientist (Analytics)
Statistical analysis, experimentation (A/B testing), causal inference. Business-facing.
ML Engineer
Building and deploying models in production. More engineering, less analysis.
Data Science Manager
Leading teams. Hiring, project prioritization, stakeholder management.
AI/ML Researcher
Cutting-edge model development. Typically requires PhD. Academia or industry labs.
MLOps / Data Engineer
Infrastructure for data and models. Pipelines, monitoring, deployment. High demand.
Go Deeper
We've talked to working professionals about every angle. Real voices, real numbers, zero sugarcoating.