YC Startup Ditches LeetCode for AI Agents
A YC-backed startup is innovating the hiring process by skipping LeetCode entirely. Instead of coding challenges, candidates build an AI agent, and the company's own agent interviews it, creating a resume-free, skills-first evaluation open to anyone.
The YC-backed startup, named Algorithmi, was founded by Lindsey Simon and Paras Garg, both of whom are former software engineers from Google's AI division. Their professional background includes work on large-scale machine learning projects, which directly informed their approach to rethinking technical assessments. The duo has been vocal about the limitations of traditional, puzzle-based coding challenges in predicting on-the-job performance. The evaluation process at Algorithmi is centered around a project-based assessment where candidates are given a specific problem and a set of tools to build a functioning AI agent. This task is designed to mirror the type of work engineers would perform in their roles, focusing on practical application rather than abstract problem-solving. The company's own AI agent then interacts with the candidate's creation, assessing its functionality, efficiency, and robustness in real-time. This new approach moves beyond the typical focus on data structures and algorithms seen in FAANG interviews. Instead, it emphasizes a candidate's ability to build and deploy a functional product. The assessment criteria include the agent's performance on a variety of tasks, the quality of the code, and the design choices made by the candidate. This provides a more holistic view of a candidate's skills compared to the pass/fail nature of many LeetCode problems. The shift to a resume-free evaluation is a key component of Algorithmi's philosophy. By focusing solely on the submitted AI agent, the company aims to reduce bias related to a candidate's educational background or previous employers. This skills-first approach is intended to open opportunities to a wider pool of talent that may not have a traditional computer science pedigree but possess strong practical engineering abilities.
