Doppel Onsite Interview for Phishing Detection Engineering Role

Company Context Doppel is a Series B startup focused on detecting and taking down social engineering attacks for brands. The Onsite Process The interview consists of a continuous 4-hour sessio

Company Context Doppel is a Series B startup focused on detecting and taking down social engineering attacks for brands. The Onsite Process The interview consists of a continuous 4-hour session structured as follows: * Introduction (15 min): Overview of the exercise. * Live Coding (45 min): Setting up boilerplate code in front of the interviewer. * Solo Coding (2.5 hours): Independent feature implementation. The interviewer leaves the call but remains available via text or email for questions. * Code Review (30 min): A walkthrough of the implemented solution and a Q&A regarding design choices. Environment and Tools The assessment is open book and open internet. However, all code must be written within Coderpad to access the necessary datasets. Candidates are permitted and encouraged to use AI tools; the company provides access to Claude Pro for the duration of the interview. Problem Statement The task involves building a "mini-Doppel" phishing detection engine. Candidates are provided with a requirements document, labeled data, and raw HTML files. The objective is to analyze these materials and programmatically identify websites that serve as "look-alikes" to legitimate brands. Solution Strategy This interview evaluates practical engineering skills rather than algorithmic theory. An effective approach involves: * Leveraging AI: Using the provided Claude Pro access to rapidly generate boilerplate code during the initial 45-minute section. * Alignment: Utilizing the live coding portion to agree on the architectural direction for the solo session. * Simplicity: designing a solution that is simple, functional, and extensible. Candidates should avoid over-engineering, as the final review requires justifying design decisions based on the requirements doc.