How Cactus AI Screened 600+ Applicants and Hired an LLM Engineer in Under a Week

Cactus AI is an early-stage AI startup. In October 2024 it needed a Senior AI Engineer with LLM expertise, had more than 600 applicants, and the co-founder running the search had no technical background to judge them. By inviting every applicant to a 15-minute Asendia AI interview, Cactus AI saved an estimated 80+ hours of interview time and filled the role in under a week.

What changed for Cactus AI

• Time to hire fell from weeks to under a week. Screening that would have taken weeks was done in days, so the role did not sit open while the founders worked through applications.
• An estimated 80+ hours of interview time saved. That is time the co-founders would otherwise have spent on first-round interviews instead of running the company.
• All 600+ applicants were invited to the AI interview, and the team spent its own interview time only on the top candidates the reports identified.
• The hire, Saif, did not score in the top 10% on coding tests. His AI interview showed a deep conceptual understanding of LLM architectures, creative problem-solving and strong communication, qualities a test-score cutoff might have screened out.

600 applicants and no technical screener

James, Cactus AI's co-founder, was handling the search himself. The role called for specialized LLM knowledge that he could not evaluate from a resume, and the company had limited resources for interviews. Speed mattered because the company was growing quickly, and so did fit: for a startup at that stage, a senior engineer also had to share the company's values and vision.

Reviewing more than 600 applications by hand and interviewing the plausible ones would have taken weeks of founder time, and a coding test alone would have ranked candidates on one narrow measure.

How Cactus AI used Asendia AI

Setup took less than a day. Cactus AI invited every applicant to create an account, upload a resume and complete a 15-minute AI interview focused on LLM work. The interviewer asked follow-up questions based on each answer, so candidates with real depth could show it. Interviews ran in parallel, which removed scheduling from the process entirely.

Within hours, the team had a transcript for each interview and an assessment of the candidate's technical proficiency and communication skills, covering both hard and soft skills. The team then limited its own interviews to the candidates at the top of that list, which is how Saif stayed in contention despite his coding test score.

Without Asendia AI, we might have missed out on one of our best-ever hires. Although his coding test scores weren't in our top 10%, Asendia's AI-driven insights revealed exceptional qualities beyond technical assessments. Our new AI Engineer, Saif, demonstrated that traditional coding tests don't fully capture what modern AI roles truly require.

James, Co-founder of Cactus AI

Ready to transform your hiring strategy? Schedule a Demo with our founders today!: https://asendia.ai/talk-to-founders