top of page
success case

With LLM HQ, 1111 Job Bank provides GenAI services for job seekers and HR.


LLM HQ is the latest BYOA product launched by LLM HQ is an AI Orchestrator to retrieve contextual prompts from enterprises for faster and safer LLM results. 


1111 Job Bank Leverages LLM HQ to recommend job vacancies based on individual resumes. 

In the Job Bank industry, most talent-to-job matching relies on personalised or rule-based recommendation systems. However, traditional AI solutions can no longer fully meet the ever-changing job skills and recruitment criteria. 

As a result, Job Banks struggle to expand their revenue amid market fluctuations, while investing in manual efforts to update and adjust resume databases. To address the rapidly changing market demands from talent and businesses, the Job Bank industry needs to explore alternative AI solutions for continuous optimization of job search efficiency for candidates and time effectiveness for employers.


​Customer Testimonial

The talent matching scenarios in job banks differ significantly from the common recommendations in e-commerce. Fortunately, with the assistance of KKLab's professional team, we were able to integrate AWS AI/ML solutions, including Amazon Comprehend, Amazon Translate, Amazon SageMaker, and Amazon Personalize. This allowed us to establish dynamic tags for resumes and job descriptions, enabling us to provide high-quality talent-matching services.

- 簡聖霖, Director of the Data Science Department, 1111 Job Bank  -

The Customer's Recommendation Chanllenges

Now, working as slashes is a new trend. If recommendations are solely based on job titles, such as “iOS Engineer,” it may only lead to finding candidates who have previously held the title of iOS Engineer, rather than identifying other professionals with more relevant experience.

Similarly, candidates may face difficulties being matched with job opportunities that align best with their experience and qualifications after completing their resumes and job requirements.

Group 4.png

How to optimize the effectiveness of job matching based on textual content when job search keywords constantly with trends?

With the rapid evolution of technology, keeping the resume data fields provided by job banks up-to-date and making real-time updates requires a substantial amount of human effort for continuous system updates and adjustments. Moreover, the existing system fields may hinder candidates or HR personnel from completing their resumes or job descriptions comprehensively, significantly limiting the precision of text-based matching.


How to overcome the limitations of traditional AI recommendations and achieve optimal talent matching when the same job title may involve significantly different recruitment criteria across other companies? addresses the changes in the job bank industry through LLM HQ by


Training a dynamic taxonomy of skill model

The customized LLM model for the job bank industry efficiently processes over 250 job skills through a dedicated natural language model tailored to the domain of job skills. It dynamically taxonomizes these skills and identifies correlations among them.


LLM model with recommendation model application for rapid matching of resumes and job descriptions

For newly registered or entry-level candidates lacking past click records, LLM can analyze the keywords in resumes and job openings. It dynamically taxonomizes their relevance and combines it with the recommendation model to perform job matching. For existing candidates, LLM can further optimize the recommended content by considering past click and interview records, using the recommendation model to weigh individual preferences.

bottom of page