How to become a Data Scientist
Data scientists turn raw data into decisions — cleaning it, modelling it statistically, building machine-learning systems and explaining what the numbers mean.
Eligibility at a glance
For India (India). Conditions marked "Not required" genuinely do not apply.
Minimum education
Bachelor's degree in statistics, mathematics, computer science, engineering or economics is the common bar; postgraduates are preferred for research-leaning roles.
School subjects
Not required
Degree / diploma
B.Tech/B.Sc/M.Sc in CS, statistics, maths or related; BCA/MCA also common
Graduation required
Not required
Minimum marks
Not required
Minimum age
Not required
Maximum age
Not required
Age relaxation
Not required
Nationality
Not required
Physical requirements
Not required
Medical requirements
Not required
Gender-specific rules
Not required
Number of attempts
Not required
- · Entry is skill-first for analyst roles; research and ML-engineering roles favour postgraduates
The path, step by step
The standard entry route. Alternative routes are listed under Education paths below.
Class 12 — Science with Maths (preferred) or any stream with strong maths
Bachelor's in statistics/CS/maths/engineering
Learn Python, statistics and SQL
Projects + Kaggle/internships
Data Analyst role (common first step)
Data Scientist / ML Engineer
What this career actually is
Data scientists combine statistics, programming and domain understanding to extract insight from data. One week may be spent building a forecasting model; the next, explaining to managers why the model's recommendation should change how the business works.
A typical day
- · Cleaning and exploring messy datasets
- · Building and validating statistical or machine-learning models
- · Writing analysis reports and dashboards for decision-makers
- · Working with engineers to put models into production
Main responsibilities
- · Frame business questions as data problems
- · Build, test and improve predictive models
- · Keep analysis honest — avoid bias, validate properly
- · Communicate findings clearly to non-technical audiences
Work environment
Office or remote, in tech firms, banks, e-commerce, healthcare, consultancies and research groups. Heavy computer work with frequent presentations.
Who this suits
- · People who like numbers, patterns and scepticism
- · Those who enjoy both coding and storytelling with data
- · Curious minds who ask 'why' behind every trend
How long the path takes
Roughly 4–6 years after Class 12 from the point the path begins.
Class 12 (with Maths)
2 years
Bachelor's
3–4 years
First analyst role / portfolio
6–12 months
Data scientist
ongoing
Timelines vary — few people follow the exact same schedule. Treat this as a map, not a schedule.
Exams on this path
Stage, pattern and eligibility details for each exam this career usually involves.
Conducted by National Testing Agency (NTA) · Twice a year in recent cycles (two sessions)
- Purpose
- Entrance test for NITs, IIITs, other centrally funded technical institutions, and the qualification screen for JEE Advanced.
- Eligibility
- Class 12 pass/appearing with Physics, Chemistry and Mathematics. Verify the current brochure for year-of-passing rules.
- Age limits
- No upper age limit in recent cycles; institutes may have their own age criteria.(may change — verify at the official source)
- Attempts
- A candidate may attempt JEE Main in multiple sessions per year; verify the current brochure for the limit on consecutive years.(may change — verify at the official source)
- Pattern & marks
- Computer-based; Mathematics, Physics, Chemistry. Negative marking applies. · 300 marks for Paper 1 in recent patterns
- Duration
- About 3 hours
- Subjects
- Mathematics, Physics, Chemistry
- Stages
- JEE Main Paper 1 (B.E./B.Tech) → JEE Main Paper 2 (B.Arch/B.Planning, for those routes) → Rank list and JoSAA/state counselling
- Notes
- Admission to IITs additionally requires JEE Advanced. Verify session dates and rules in the current brochure.
Conducted by IISc and IITs (organising institute rotates) · Once a year
- Purpose
- Admission to M.Tech/MS and PhD programmes, and recruitment by public sector undertakings (PSUs) and research organisations.
- Eligibility
- Bachelor's degree holders in engineering/technology/architecture, or master's in relevant science streams; candidates in the final year may apply per the current brochure.
- Age limits
- No age limit.(may change — verify at the official source)
- Attempts
- No attempt limit.(may change — verify at the official source)
- Pattern & marks
- Computer-based; discipline paper plus a general aptitude section. Question types include MCQ, multiple-select and numerical answer. · 100 marks (long-standing scheme)
- Duration
- 3 hours
- Subjects
- Engineering mathematics, Chosen discipline's core subjects, General aptitude
- Stages
- Single computer-based paper in the chosen discipline → Score used by institutions/PSUs for their own processes
- Notes
- Each PSU runs its own recruitment on GATE scores; cutoffs vary by organisation and year.
Conducted by National Testing Agency (NTA) · Once a year
- Purpose
- Admission to undergraduate (and postgraduate, at many universities) programmes at central and participating universities.
- Eligibility
- Class 12 pass/appearing for UG programmes, subject to each university's own programme requirements.
- Age limits
- No uniform age limit; university-specific rules may apply.(may change — verify at the official source)
- Attempts
- Not applicable.(may change — verify at the official source)
- Pattern & marks
- Subject-wise computer-based tests: a language paper, domain subjects and a general test, chosen per the programme's requirement map. · Varies by papers chosen — verify the current notice.
- Duration
- About 45–60 minutes per paper (recent cycles)
- Subjects
- Languages, Domain-specific subjects, General test
- Stages
- CUET computer-based test → University counselling and admission
- Notes
- Each university maps its programmes to required CUET papers differently — check the specific university's admission bulletin.
What to study — complete checklist
Tick topics as you learn them. Your ticks are saved in this browser.
Skills required
Qualifications get you in; these keep you there.
Technical skills
- Python (pandas, NumPy, scikit-learn)
- SQL
- Statistics and experiment design
- One ML framework
Practical skills
- Cleaning real, messy datasets
- Designing A/B tests
- Turning vague questions into measurable ones
Communication skills
- Explaining models without jargon
- Visual storytelling
- Writing crisp analysis memos
Tools & software
- Jupyter notebooks
- Tableau or Power BI
- Git
- A cloud ML platform
Languages
- Python is the default
- R in statistics-heavy teams
- SQL everywhere
Certifications
- Optional — respected ones exist but projects matter more
Portfolio
2–4 end-to-end projects with clear write-ups of the problem, method and result
Education paths
Every major route into this career — including routes that skip a degree where they exist.
Degree route
Typical duration: 3–4 years- Class 12 with Maths
- B.Sc/B.Tech in stats, CS or maths
- M.Sc/M.Tech for research-leaning roles
- Analyst → scientist progression
Analyst-transition route
Typical duration: 2–3 years- Any quantitative degree
- First job as data analyst
- Self-study ML, build projects
- Move into data scientist role
Skill route (starter roles)
No degree neededTypical duration: 1–2 years- Strong maths self-study
- Python + SQL mastery
- Portfolio of real-data projects
- Junior analyst roles at smaller firms
Is college compulsory? No — recognised non-college routes exist (shown above).
Without a degree: Junior analyst roles at smaller firms hire on demonstrated skill; a public portfolio of real-data projects is the entry ticket. Research and senior roles generally require degrees.
Entrance exams: JEE/GATE for engineering and MTech routes, University-specific tests for B.Sc data science
Career progression
How roles typically advance. Alternative branches shown where the path forks.
- Data Analyst
- Data Scientist
- Senior Data Scientist
- Lead / Principal Data Scientist
- Head of Data Science
Where the jobs are
Earning potential
Broad ranges only — pay is never guaranteed and varies hugely by employer, city and seniority.
| Entry level | Broadly ₹4–10 lakh per year for analysts and junior scientists |
|---|---|
| Mid-career | Broadly ₹12–30 lakh per year |
| Experienced | ₹30 lakh+ per year at senior/specialist levels |
Highly variable by industry and city; ranges are observations, not promises.
What you are up against
An honest picture of competition and effort — no fake difficulty scores.
Competition
High — many applicants, but genuinely strong portfolios are rare and stand out
Study load
Heavy at entry: maths, statistics, coding and domain sense together
Selectivity
Top research labs are extremely selective; business analytics roles are more accessible
Major challenges
- · Interviews test maths and coding both
- · Real data is messier than tutorials
Time commitment
6–12 months of focused preparation for the first role
What candidates commonly struggle with
- · Statistics intuition
- · Communicating results to non-technical people
Documents you will need
Commonly required for applications and admissions. Exact requirements are set by each authority — always check the official notice.
- ✓ Aadhaar card or other government photo ID
- ✓ Class 10 and Class 12 marksheets and certificates
- ✓ Recent passport-size photographs
- ✓ Scanned signature (as per the notified size and format)
Recommended Books
Recommended Books
Book recommendations are being curated.
Checked for Data Scientist — nothing is listed until it is genuinely recommended.
Frequently asked questions
Do I need a PhD?
Not for most industry roles. A PhD matters mainly for research scientist positions at labs and for very specialised modelling work.
Data analyst vs data scientist — what's the difference?
Analysts describe and explain what happened using queries and dashboards; scientists build predictive models on top of that foundation. Analyst is the common first step.
Which degree is best?
Statistics, mathematics, computer science and engineering all work. What matters most is strong maths plus coding proof.
Is this career oversaturated?
Entry-level applicant numbers are large, but hiring managers report a shortage of candidates who can actually clean data, model honestly and communicate. Skill depth beats certificates.
Can I enter this career after Class 10?
Not directly. The standard entry path starts later — see the roadmap above for where it actually begins.
Can I enter this career after Class 12?
Yes — Class 12 is a genuine entry point for this career. Look at the education routes and roadmap for the exact next step.
Which stream should I choose in Class 11–12?
No single stream is mandatory. Pick the stream that keeps the subjects you will need — the eligibility section lists what matters for this career.
Is a degree compulsory?
No — this career can be entered without a degree. Junior analyst roles at smaller firms hire on demonstrated skill; a public portfolio of real-data projects is the entry ticket. Research and senior roles generally require degrees.
What is the age limit?
There is no fixed age limit for the standard entry path. Where a specific exam or employer applies one, it is listed in that exam's or employer's own rules — always verify from the official source.
Is there an entrance exam?
Yes — this career commonly involves entrance exams. See the Exams section for the stages, pattern and official links.
How long does it usually take?
Roughly 4–6 years after Class 12, counting from the point where the path begins. Timelines vary — people repeat years, switch routes and start at different points, so treat this as a map, not a schedule.