Natoora Data Analyst Interview Experience — Grilled on ETL and LLM Claims After Admitting No Pricing Background

The interviewer cared a lot about how real and grounded the projects on my resume were, especially the logic behind the data flow (Data Infrastructure). If you have flashy words like LangChain or LLM on your resume, you need to be ready to get asked "why didn't you just use plain SQL/Python?"

Interview Process Recap

The interviewer first confirmed whether I had pricing experience — I said straight up that I didn't, so they pivoted to testing my data-processing ability and logical thinking instead. They asked very detailed questions about the ETL pipeline and LLM project on my resume. My advice: you need to understand every single keyword on your resume down to the underlying-architecture level.

Core Questions, Compiled (All of Them)

Part 1: Background & Tooling

• (Direct Opening) I see you don't really have any pricing experience. Is that correct?
• Do you have any experience in financial analysis or related fields, even if it's academic?
• You use SQL/Python/R, but most of our work is on Google Sheets. How do you feel about that?
• Can you give me a sense of the most complex model that you've built or maintained on Google Sheets?
• What technical infrastructures do you feel most comfortable working in, whether academic or professional?

Part 2: Resume Project Deep-Dive (LLM/Automation)

• You built an automated pipeline using LangChain and LLM to generate daily summary reports. What was the output specifically and where did it go?
• What was your data source to begin with?
• If it was a CSV with core metrics, why would you necessarily need LangChain and LLM to generate daily summary reports?

Part 3: Resume Project Deep-Dive (ETL/Data Engineering)

• You developed an ETL pipeline to pre-process 25,000 CSV files. How exactly were you receiving the data?
• Where were you extracting it from? Is it just that someone would give you a CSV file and you process it and post it to the database?
• How was it an automated ETL pipeline specifically? (The interviewer kept pressing on manual vs. semi-automated vs. fully-automated.)
• When was the transform triggered? (They were testing concepts like cron jobs, event-driven triggers, etc.)
• What was the infrastructure for the centralized database? Was it a hospital database, local, or cloud? Was it ever put into production?

Part 4: Questions I Asked the Interviewer

• How do you expect a candidate to show "exact experience"?
  Interviewer's answer: the key is understanding trade-offs, being able to clearly articulate your decision points, and staying flexible when working with different tools and unstructured data.
• Infrastructure is the main focus: even for an internship project, you need to be clear on how the data actually flows (from source, to server, to database).