In today’s rapidly evolving job market, technology is reshaping every aspect of the hiring process. Did you know that 42% of companies are now using AI to assess job applications? This staggering statistic highlights the growing influence of artificial intelligence in determining who gets hired.
But what does this mean for job seekers? How do these AI systems evaluate resumes, and what are they really looking for? To shed light on this increasingly important topic, we decided to turn the tables and let AI analyse some sample resumes.
We wanted to help humans understand what AI really thinks when it looks at a CV. So, we used an advanced resume chatbot to review three different quality levels of resumes: moderate, very good, and perfect.
When you submit your resume, the first thing an AI notices is its format. Clean, well-structured documents are easier for AI to parse and understand. Think of it as making a good first impression, but with pixels instead of a handshake.
AI Perspective: A well-formatted document allows efficient extraction and categorisation information. Inconsistent formatting or unconventional layouts can lead to errors in my interpretation, potentially causing me to misclassify or overlook important details. I appreciate resumes that use standard headings, consistent font sizes, and clear delineation between sections.
While humans might skim a resume, AI systems can analyze every word in microseconds. They’re typically programmed to look for specific keywords related to the job description, as well as indicators of experience and skills.
Generic objective statements are the bane of both human and AI recruiters. An AI system is likely to give more weight to objectives that closely match the job description and company values.
That shouldn’t come as a surprise, given the underlying technology behind LLMs operates via statistics. Instead of saying you “improved sales,” stating that you “increased sales by 30% over six months” gives the AI concrete data to work with.
For example, when a resume states ‘Led a team of 15 and reduced project delivery time by 25%,’ it provides the AI with clear, measurable indicators of success. This allows it to more accurately assess a candidate’s potential impact. Vague statements like ‘Responsible for team leadership’ are less informative.
We’ve presented the AI with three different CVs. First, we’ll give you the AI’s analysis, before giving you a human recruiter’s perspective on that analysis.
John Doe’s CV presents several issues that hinder its effectiveness:
Overall, this CV doesn’t provide enough specific, relevant information to make a strong case for John’s candidacy for most professional roles.
The AI’s assessment of John Doe’s CV is generally accurate but perhaps overly harsh. While our human recruiter would likely agree with many of the points raised, they might also:
However, the AI correctly identifies the lack of specificity as a major weakness. We would indeed prefer more concrete examples of achievements and relevant skills.
The AI’s focus on formatting is interesting. While humans do appreciate well-formatted CVs, they’re generally more forgiving of basic layouts, especially for junior positions.
Jane Smith’s CV shows significant improvements:
This CV effectively communicates Jane’s qualifications and achievements in digital marketing. It provides enough specific information to make a strong case for her candidacy in related roles.
The AI’s positive reception of Jane Smith’s CV aligns closely with how our human recruiter would perceive it. The analysis correctly highlights:
We agree with all these points. However, a human might also:
The AI’s appreciation for clear formatting mirrors human preferences, though a human might not consciously analyse this aspect as thoroughly.
Michael Johnson’s CV exemplifies an optimal presentation of professional qualifications:
This CV provides comprehensive, relevant, and quantifiable information about Michael’s qualifications and achievements. It makes a very strong case for his candidacy in project management roles, providing multiple data points to assess his potential performance.
The AI’s glowing review of Michael Johnson’s CV is largely in line with how a senior human recruiter would evaluate it. The analysis correctly emphasises:
We would likely be equally impressed. However, there are a few nuances a human recruiter might consider that the AI doesn’t mention:
The AI’s methodical breakdown of each section is more systematic than a typical human review, which might focus more on overall impression and standout elements.
Our analysis reveals some key insights into how AI evaluates CVs compared to human recruiters:
AI systems prioritise quantifiable achievements and specific information, while also placing high importance on formatting and structure. They offer consistent, thorough analysis across all applications, but may miss nuances that humans pick up on, especially for less experienced candidates.
Human recruiters, on the other hand, often read between the lines and consider factors like cultural fit and personality. They may be more forgiving of less polished CVs, particularly from junior applicants.
For job seekers, the challenge lies in crafting a CV that appeals to both AI and human reviewers. This means highlighting specific achievements and skills while also conveying personality and potential cultural fit.