There's no shortage of data science courses, tutorials and learning materials available online. However, it can often feel intimidating as a beginner to figure out where to start. This short & concise guide is meant to help you build your data science career the right way and answer the most common doubts you may have: which topics you should learn, in what order, which projects you should/shouldn't do, what you should put in your resume etc. It would help to bookmark this page and refer to it from time to time, whenever you need some guidance or feel lost.

1. Skills and Domains
2. Courses, Certifications & Books
3. Projects
4. Job roles
5. Resume & Professional profile
6. Job Hunting & Interviews

All the above ingredients are important for getting started with a career in data science, so don't ignore any of them. Depending on where you currently stand, it might take anywhere from 3-4 months to a year to build a strong profile. But don't worry, you'll have a lot of fun along the way.

• Domains: Wrangling, EDA, Visualization, Machine Learning, Deep Learning
• Coding: Python, R, Frameworks (SKLearn, TensorFlow, PyTorch etc.)
• Math: Linear Algebra, Statistics, Basic Calculus
• Misc.: Experimentation, Documentation, Presentation, Storytelling, Excel
• Deployment (bonus): Backend, UI, Cloud (AWS/Heroku), DevOps

Data science lies at the intersection of coding, mathematics and scientific experimentation, which makes it very interesting but also quite challenging to master. Make sure you are comfortable with as many of the domains & skills as possible. As a data science practitioner you will often work closely with product managers, software developers, company executives and other stakeholders, so documentation, presentation and storytelling skills are also important.

• Just pick 2-3 good ones and cover all relevant topics
• Practical: Coursera ML, FastAI, mlcourse.ai
• Theory: Deelearning.ai, CS231n, Deep Learning Book
• Coding: Python for DS & ML (Udemy), Deep Learning with Python (Manning)
• Certifications are nice-to-have, if you can afford them, but not compulsory

Most academic programs don't offer many courses in data science, as it's a new and evolving field. But there are several great courses online, and you can start learning in your free time by spending 1-2 hours a day. Do enough courses so that you feel confident with the material, but don't go overboard doing dozens of courses. It's more important to work on projects to improve your practical skills.

• Projects matter much more than grades, courses or certifications
• Course assignments and standard datasets DON’T count
• Find interesting, unique or personal datasets & problems (from Kaggle, UCI, more)
• E.g.: Browing habits, Movie Genre classifier, WhatsApp insights, Survey analysis, more
• Focus on code quality, presentation & documentation

Working on projects is a great way to learn, and it's also the best way to showcase your skills to potential employers. Make sure your projects are of high quality and represent your best work. Trivial projects like Titanic & MNIST classification don't count, as everyone does them. Make your projects unique, personal and creative. You never know, an intersting project could land you a data science job, without even trying!

• Business Analyst: Gather insights & prepare reports using GUI-based tools (Excel, Tableau etc.)
• Data Analyst: Analyze company data and create reports using code (R, Python, Jupyter etc.)
• ML Engineer: Implement, test, deploy and monitor ML models from research into production
• Data Engineer: Set up jobs & pipelines for big data aggregation, analysis & reporting
• Data Scientist: Research, train, evaluate and improve ML models to solve business use cases

Spend some time understanding the responsibilities for each kind of role, and decide which one you want to target. Talk to others working in these roles to get a sense of their day-to-day work, and understand whether you have the right skills for it. Note that these roles are farily flexible, so you can start with any one of them and switch to other ones within a year or two.

• Share your projects, Jupyter notebooks & demo videos online (GitHub, LinkedIn, Jovian)
• Write blog posts and tutorials (on Medium, Hackernoon etc.)
• Create presentations and working demos online (web applications)
• Contribute to the community (forums, meetups, presentations)
• Highlight unique projects & include links to the above in your Resume

A good Resume is one that provides ample evidence for every project/skills listed within it. If you've followed all the previous steps, you should have several projects, blogs posts, presentations etc. posted online that you can list and link to from your Resume. Participating in community events and forums is a great way to meet potential employers or other professionals who can refer you for a job.

• Rather than finding jobs, it’s better to be found (via your projects, blogs or professional network)
• Practice algorithms & data structures (GeeksForGeeks, LeetCode)
• Have a solid understanding of the basics (math, coding, libraries)
• Don't lose heart. Be patient and learn from every rejection. Ask for feedback.
• Keep working hard. Never stop improving your skills & building your profile.

Remember, don't jump to job hunting straightaway. It's 2020, and with the multitude of resources available online, nobody's going to hire or train you if you don't already have a solid foundation in the domain. If you put in the hardwork for a few months and build a strong profile, you'll have a much easier time getting interviews & job offers. Keep at it, and don't forget to enjoy the journey!

Thanks for reading! If you found this guide useful, please share it and spread the word. For more data science content join our Telegram channel: t.me/jovian.ml