News: Telta Wins Two Tasks at TalentCLEF, a Global HR AI Challenge

Picture a hiring process with hundreds of resumes piling up. Finding the person who best fits the role usually comes down to a recruiter’s time and intuition. The less time recruiters have, the easier it is to miss the right candidate; the more they rely on intuition, the less consistent their evaluations become.

In the end, hiring accuracy depends on two questions.

1. How accurately can you read a resume?
2. And how precisely can you define the competencies a role truly requires?

Telta’s HR AI solution took first place in both core tasks at TalentCLEF, a global HR AI challenge that evaluates exactly these two areas.

‍

TalentCLEF: A European Academic Challenge That Tests AI in Real HR Settings

TalentCLEF is an HR-focused AI challenge run under CLEF, a European academic evaluation forum for natural language processing (NLP) and information retrieval. Unlike general AI benchmarks that rank models purely on academic dataset scores, it evaluates AI performance on data and tasks drawn from real-world HR, which is what makes it meaningful.

This year’s two core tasks map directly onto problems recruiters face every day.

• Job–Resume Matching Analyze job postings and candidate resumes to recommend the candidates with the strongest job fit. Evaluation also covered a multilingual setting in English and Spanish.
• Job Skill Classification Extract the competencies a specific role requires and classify them as required or preferred.

Reading resumes quickly and pinpointing the competencies a role truly needs: these are where recruiters spend the most time, and where accuracy translates directly into hiring quality.

‍

Related Reading

• AI-Era Talent Strategy: Filling the Skills Data Gap in Your HCM and HRIS
• A 4-Step Guide to Becoming a Skills-Based Organization

‍

Telta Took First Place in Both Tasks: Job–Resume Matching and Job Skill Classification

Telta ranked first in both tasks. In job–resume matching, it maintained high performance even in a multilingual setting. In job skill classification, it was also evaluated on how precisely it distinguished required from preferred competencies.

“This result is significant because it validates, on a global stage, AI technology that can be put to work in real HR settings,” said Dr. Soyoung Jeon, head of Telta. “We’ll keep focusing on advancing our HR AI so companies can find talent and assess competencies faster and more precisely.”

Finding the person who best fits a role among countless resumes, and defining the competencies a role truly requires: Telta’s HR AI has now been validated as world-class in both.

Successful hiring starts with accuracy at the very first step. Telta will keep strengthening its HR AI so that, from that starting point, companies can meet candidates on the basis of data.

Money Today, “‘The AI That Reads Resumes Best’: Classum’s Telta Takes Two Titles on the Global HR AI Stage” (in Korean)
Read the original: https://news.mt.co.kr/