Data engineering roadmap and certifications

You have now seen the whole data engineering stack in this track. This lesson turns it into a plan: what to learn in what order, which projects prove it, which certifications are worth considering, and how interviews for data engineering roles usually work.

Everything about job markets and exams changes, so treat this as a map, not a promise. No lesson can promise a job or a salary. What it can do is show you a sensible order of learning and the evidence employers look for.

Titles differ between companies. A "data engineer" at a startup may do everything from dashboards to servers, while at a large company it may mean one narrow part of a big platform. Read the job description, not just the title.

This order works for most people. Times are rough guides for studying a few hours a day, and you will move at your own speed.

A project is evidence. Employers cannot see your skills, but they can read a repository. Strong portfolio projects share these traits:

- They solve a clear problem, written in the README in two sentences. - They use realistic messy data (public datasets, an API, or data you generate with planted problems) instead of a perfect CSV. - They include tests, quality checks and a way to re-run safely. - They have an architecture diagram and instructions that work in five minutes. - They show decisions: why this tool, what you would change at 100 times the size.

Ideas: the end-to-end Brew & Bean pipeline from this track extended with Airflow and dbt; a pipeline that pulls a public API such as weather or cricket scores daily into a warehouse; a small Kafka stream of simulated orders feeding a live dashboard; a RAG ingestion pipeline over your college's public documents with incremental updates. Two or three finished projects are worth more than ten half-done ones.

Certifications help most when you are a fresher without work experience or when you are moving to a new cloud. They do not replace projects. Most have exam fees and rules that change, so always check the provider's official page for the current exam name, price and syllabus. These are the ones commonly asked about, as of the time of writing (2026).

Processes differ, but a data engineer hiring loop commonly includes:

- Online test or screening. SQL queries (joins, window functions, aggregations) and sometimes Python or basic data structures. - SQL round. Live queries: top N per group, running totals, deduplication, finding gaps, explaining a query plan. - Python or coding round. Parse a file, transform records, handle bad data. Often moderate difficulty, not hard competitive programming. - Data modelling and system design. "Design a pipeline for order data" or "model a food delivery app for analytics." You draw sources, ingestion, storage, transformation, orchestration, quality and serving, and discuss trade-offs. - Project deep dive. Detailed questions about your own project. This is where real projects separate from copied ones. - Behavioural round. Handling a pipeline failure, working with analysts, explaining a technical decision simply.

You are an interviewer for a data engineer role at [type of company, e.g. an e-commerce startup in Bengaluru]. I am a [fresher / 2 years experience] candidate. Run a mock interview with one question at a time and wait for my answer before continuing.

Rounds: 1) three SQL questions of increasing difficulty, 2) one data modelling question, 3) one pipeline design question, 4) two behavioural questions.

After each of my answers, give brief feedback: what was right, what was missing, and a better answer in two or three sentences. At the end, give me a score out of 10 for each round and three things to practise.

What to expect: Good for practice, but check any SQL or technical claims it makes. You can also practise with the InterviewAI mock interview feature.

If you are a student, add internships, final-year projects and campus placement prep. If you are a developer or analyst switching careers, start from the projects stage and use your existing domain knowledge, which employers value, in your portfolio.

AI coding assistants now write a lot of SQL and boilerplate pipeline code. That makes fundamentals more valuable, not less: you must read generated code, spot wrong joins and unsafe deletes, test it, and know when its answer is plausible nonsense. The parts AI does not do for you are the parts of the job that matter most: understanding the business question, designing reliable systems, choosing trade-offs, handling failures at 2 am, and governing data safely. Use AI as a fast junior who needs review (see *What data engineers do, and how AI changed the job*), and keep building judgement.