Data Engineering Intern (2026 Batch) - Poshmark Chennai

Let’s be real for a second. Being a 2026 graduate (currently in your 3rd year) is a weird place to be. You are watching your seniors struggle with placements, you are hearing rumors about "recession," and you are wondering if learning React or Java is actually going to get you a job.

You want a job that is "AI-proof," right? You want a job that pays incredible money but doesn't require you to be a competitive coding wizard who solves Hard LeetCode problems in 5 minutes.

I have the answer for you: Data Engineering.

Data Engineering is the backbone of the modern internet. AI doesn't work without data. Analytics doesn't work without data. And Poshmark—one of the biggest social commerce fashion marketplaces in the world—is looking for you to help build that backbone.

They aren't looking for a senior expert. They are looking for an Intern. A 2026 graduate. Someone they can mold, teach, and eventually hire full-time. This isn't just an internship; this is a golden ticket to skipping the "fresher struggle" completely. If you want to work on systems that handle Petabytes of data (that’s millions of Gigabytes) and learn tools like Spark, Kafka, and AWS before you even graduate, this is the role you need to fight for.

1. Why This Job is an Amazing Opportunity

✅ Benefit 1: You Will Touch "Big Data" for Real

A lot of internships claim to be "Big Data," but they just give you a 50MB Excel sheet and tell you to make a pie chart. Poshmark is different. The job description explicitly mentions handling "terabytes to petabytes" of data. This is rare. In college, you handle small arrays. Here, you will write code that processes millions of transactions, user clicks, and image data every single hour. Experience with this scale is what separates a ₹4 LPA engineer from a ₹20 LPA engineer. Having "Processed Petabyte-scale data using Spark" on your resume will make you unstoppable.

✅ Benefit 2: The "Modern Data Stack" Education

Look at the tools listed: Databricks, Airflow, Spark, AWS, Kafka. This is the "Avengers" team of data tools. Most companies are still using old, clunky databases. Poshmark is using the modern cloud stack. If you tried to learn these tools on your own, it would cost you thousands of dollars in cloud bills. Poshmark will pay you to learn them. You are getting a world-class education in Data Engineering that is better than any Master's degree.

✅ Benefit 3: A Clear Path to a PPO (Pre-Placement Offer)

Companies rarely hire interns for "critical" infrastructure roles unless they plan to keep them. They are investing time to teach you their complex systems. If you do well in this internship, the chances of converting this into a Full-Time Offer for 2026 are incredibly high. Imagine entering your final year of college with a high-paying job offer already in your pocket. No stress, no placement anxiety. Just you, your degree, and a fat offer letter.

2. Role Details

3. The "What, How, & Why" of This Role

What You Will Actually Do:

You are the Plumber of the Internet.

[Image: A messy pipe system turning into a clean, organized flow]

Imagine Poshmark is a giant mall. Every time a user clicks "Like" on a dress, or searches for "Nike Shoes," or buys a handbag, that is a piece of data.

Right now, that data is messy. It's raw. It's scattered everywhere.

Your Job: You build the pipelines (using code) to grab that data, clean it up (remove duplicates, fix errors), and store it neatly in a "Data Warehouse" so the Data Scientists can use it to build recommendation algorithms (e.g., "Because you liked this dress, you might like these shoes").

How You Can Succeed in the First 90 Days:

• Month 1 (The Student): You will feel stupid. That is normal. Your goal is to understand the "Schema." What tables exist? What does the data look like? You will learn to write basic Python scripts to move files from AWS S3 to a database.

• Month 2 (The Builder): You will be given a ticket. "Hey, the marketing team needs data on user signups from Instagram." You will write a Spark job to fetch that data, transform it, and load it into Redshift. You will use Airflow to schedule it to run every morning at 6 AM.

• Month 3 (The Owner): You will optimize. "This query takes 2 hours to run. Can you make it run in 20 minutes?" You will tweak the code, partition the data, and make it fly.

Why This Role is a Stepping Stone:

Data Engineering is currently the most "in-demand" tech role, even more than Data Science. Why? Because you need 5 Data Engineers to clean the data for every 1 Data Scientist who analyzes it. By starting this career in 2026, you are positioning yourself in a market where there is a massive shortage of talent. You will command high salaries and have job security for decades.

4. Interview Preparation Guide (With Master Class Resources)

Poshmark is a product company. They will grill you on fundamentals. Do not memorize; understand.

Where to Practice:

• Coding: LeetCode. Filter by "Python" and "Easy/Medium." Focus on Arrays, Strings, and HashMaps. You don't need intense Dynamic Programming, but you need clean logic.

• SQL: HackerRank or LeetCode Database Section. You must know Joins and Group By.

5. Key Concepts to Revise (Deep Syllabus)

Concept 1: ETL vs. ELT (The Foundation)

Focus: Transformation logic timing (Pre-load vs. Post-load) and Warehouse architecture

Master Video: ETL vs ELT Explained - IBM Technology

About: This is the fundamental design choice for any pipeline. You must explain that ETL (Extract-Transform-Load) is for strict, compliance-heavy data, while ELT (Extract-Load-Transform) allows for faster "Raw Data" loading into the lake, giving you flexibility to transform it later using powerful cloud warehouses.

Concept 2: Apache Spark (The Engine)

Focus: Distributed Computing, RDDs vs. Dataframes, and In-Memory Processing

Master Video: Learn Apache Spark in 10 Minutes - Darshil Parmar

About: Poshmark deals with massive fashion inventory data. You need to understand how Spark moves beyond Excel's limits by splitting a 1-billion-row file into small "Partitions" and processing them across multiple server nodes simultaneously.

Concept 3: SQL Window Functions (The Test)

Focus: Analytical functions (RANK, LEAD, LAG) for row-relative calculations

Master Video: SQL Coding Interview Question Using A Window Function - StrataScratch

About: Standard GROUP BY collapses rows, but Window Functions keep the rows while adding a calculation. This is critical for questions like "Calculate the running total of sales" or "Find the difference in price between this item and the previous item."

Concept 4: AWS S3 & Data Lakes

Focus: Bucket structure, Object Storage, and File Formats (Parquet/JSON)

Master Video: AWS S3 Tutorial For Beginners - Simplilearn

About: You need to treat S3 not just as a hard drive, but as an API. Understand the concept of "Buckets" (folders) and why Data Engineers prefer "Parquet" files (columnar storage) over CSVs for faster querying and lower costs.

Concept 5: Airflow (The Scheduler)

Focus: DAGs (Directed Acyclic Graphs), Task Dependencies, and Backfilling

Master Video: What is Apache Airflow? For beginners - Data with Marc

About: Automation is key. You must describe how Airflow uses Python code to define a "DAG"—a visual flow where Task B waits for Task A to finish. This ensures that your daily sales report doesn't run until the sales data has actually arrived.

Concept 6: Kafka (Real-Time)

Focus: Topics, Partitions, Producers, and Consumers

Master Video: Getting to Grips with Confluent: Kafka Basics - Somerford Associates

About: For features like "Real-time Bidding" or "Fraud Detection," batch processing is too slow. Kafka acts as a high-speed message bus, allowing you to react to events (like a user clicking "Buy") the millisecond they happen.

Real-World Interview Questions:

• ❓ Coding (Python): "Given a log file with millions of lines, how would you find the top 5 most frequent error messages? (Hint: Use a HashMap/Dictionary)."

• ❓ SQL: "Write a query to find users who purchased an item yesterday but did not purchase anything today. (Hint: Left Join or NOT IN)."

• ❓ Conceptual: "What is the difference between a Data Lake and a Data Warehouse?" (Answer: Lake = Raw/Messy, Warehouse = Clean/Structured).

• ❓ Big Data: "Why do we use Parquet file format instead of CSV for big data?" (Answer: Parquet is columnar and compressed, making it much faster to read).

• ❓ Scenario: "A data pipeline failed at 3 AM. How would you debug it?" (Answer: Check logs, check data arrival, check schema changes).

• ❓ Behavioral: "Tell me about a time you had to learn a new technology quickly for a project."

6. Why Join Poshmark?

It’s Not Just "Shopping," It’s Social Media.

Poshmark isn't like Amazon where you just buy and leave. It’s a community. People follow each other, share closets, and attend virtual "Posh Parties." This means the data is complex. It’s a mix of transaction data (money) and social graph data (likes/follows). Working on this "Social Commerce" intersection is intellectually challenging and exciting.

A Culture of "Thriving"

The job description starts by saying "Confidence can sometimes hold us back... please apply." This is huge. Most companies write intimidating JDs. Poshmark is explicitly telling you: We value potential over perfection. They are looking for people who are "eager to learn." This signals a supportive, mentorship-heavy environment where it is okay to ask questions. For an intern, this is the most important factor.

Chennai’s Growing Tech Hub

The Poshmark office in Chennai is a key R&D center. You aren't working in a satellite support office; you are working where the core engineering happens. You will likely be collaborating directly with teams in the US (Redwood City), giving you global exposure.

7. FAQs

• Q: I am a 2025 graduate. Can I apply?

  • A: NO. The title explicitly says "2026 Graduates Only." If you apply as a 2025 grad, your resume will likely be auto-rejected by the system.

• Q: I don't know Scala. Is that okay?

  • A: Yes! The JD says "Python/Scala preferred." If you are strong in Python, that is enough. You can learn Scala on the job if needed.

• Q: Do I need to be a Cloud Expert?

  • A: No. "Exposure" or "Academic knowledge" is enough. If you have watched a few YouTube videos on AWS S3 and understand what it is, you are qualified to apply.

• Q: What projects should I put on my resume?

  • A: Don't put "Library Management System." Build a "Stock Price Analyzer" or "Twitter Sentiment Analysis" pipeline. Something that takes data, processes it, and shows a result.

8. Final CTA & Important Links

🔥 Urgent Notice: This role was posted 24 days ago. In the intern world, that is a long time. Positions might close any minute. Stop thinking and APPLY NOW.

APPLY ON COMPANY SITE