AI Research Engineer (Early Career)

Executive Summary

The candidate Jovanović is a promising early-career engineer with a strong academic pedigree in CS and mathematics and demonstrated ability to contribute in professional software engineering settings. However, they are a poor fit for this specific role because they lack any demonstrated experience with LLMs, RAG pipelines, AI agents, or the ML ecosystem that underpins the entire job description. Their technical profile is that of a systems/software engineer in transition, not an applied ML engineer. The application itself — an empty cover letter and no code sample — further reduces confidence in their motivation for this role specifically. They would be a more competitive candidate for a general software engineering or data engineering role, or for this role in 12–18 months if they actively pursues ML/AI projects during their Master's program.

Assessment Reasoning

The NOT_FIT decision is driven primarily by a fundamental skills gap: the candidate meets fewer than 25% of the required technical skills for this role. The position explicitly requires hands-on experience with LLMs, RAG, AI agents, LangGraph, vector databases, and prompt engineering — none of which appear anywhere in their resume, projects, or online profiles. While their C++ and Python skills are noted and their academic background is strong, the job description explicitly states it is not a notebook role and requires production-level AI/ML engineering. Their only Python experience is in environmental data science (satellite imagery processing), which is orthogonal to the LLM/agent stack. Additionally, the low-effort application (no cover letter content, no code sample, no GitHub) reduces confidence that they are genuinely pursuing applied AI work. They are below the BORDERLINE threshold and does not warrant an HR review at this time.

Interview Focus Areas

No direct code sample was submitted, making definitive code quality assessment impossible. The GitLab links in the resume point to coursework projects, which suggest a competent junior developer with multi-language exposure. There is no evidence of ML, data pipeline, or LLM-related code, which is the most relevant dimension for this role.

The candidate has a solid academic foundation in CS and mathematics and has shown they can operate in real engineering environments (Wärtsilä internship with C++/Qt5). However, their profile has no meaningful overlap with the core technical requirements of this role — there is no evidence of LLM, RAG, or AI agent work, and their ML exposure appears limited to environmental data science tooling. They are not yet positioned for an Applied ML/LLM engineering role.