-Legal Entity: Hyperconnect -Brand: Match Group AI
About Match Group AI
Match Group AI is a central technology organization that solves core problems in online dating with AI. We focus on discovering challenges in key areas of user experience such as profiles, matching, and Trust & Safety. We innovate user experience by solving these challenges through cutting-edge AI technology and data-driven approaches. Furthermore, we expand the group's common technology foundation through collaboration with Match Group's global dating brands such as Tinder and Hinge.
If you'd like to learn more about the projects our team is working on, please check out the following post: ๐ "Introducing the Match Group AI Team"
About Match Group AI ML Team
The ML Team is composed of ML Engineers who apply AI/ML technology to various Match Group services. The team's work begins with identifying and defining problems that arise in the process of creating and operating actual products. We develop or reproduce State-of-the-Art (SotA) models best suited to solve problems, and deploy the completed models stably and efficiently in mobile and server environments. Subsequently, we build the service's AI Flywheel through continuous monitoring and improvement. In this process, we collaborate closely with various specialized organizations such as backend/frontend/DevOps engineers, data analysts, and PMs to create AI experiences that have real impact on actual users. For more details about how we work, please refer to the following:
- [How AI Lab Works] Head of AI - Shurain Interview
- AI in Social Discovery(Blending Research and Production)
Research and Publications
Some of our work achievements are published externally as papers or open-source code. When building ML models for product use, existing research is often insufficient. To fill these gaps, project participants collaborate to organize the meaningful parts of the research results and, when possible, publish them along with code. As a result, we have achieved approximately 20 external research accomplishments, including the following:
- 2024 CUPID: Real-time Session-based Mutual Recommendation System for 1:1 Social Discovery Platforms ICDM Workshop Presentation
- 2023 TiDAL: Active Learning Technique Based on Model Behavior in Efficient Learning Process ICCV 2023 Publication
- 2023 Research on Setting Thresholds to Simultaneously Satisfy Multiple Classification Criteria in Moderation Environments WSDM 2023 Publication
- 2022 Research on Enhancing Semantic Diversity in Dialogue Generation EMNLP 2022 Publication
- 2022 Effective Learning Method in Environments with Severe Label Noise ECCV 2022 Publication
- 2022 Chatbot Research on Imitating Target Characters Using Only a Few Utterances NAACL 2022 Publication
- 2022 Research on Improving Performance Using Examples in Dialogue Generation Models ACL 2022 Workshop Presentation
- 2022 Distillation Technology Research for Audio Classification in Mobile Environments ICASSP Publication
Infrastructure
For ML research to proceed well, infrastructure for deep learning training must also be well-equipped. We maintain our own deep learning cluster (160 A100 GPUs and 40 H100 GPUs in total) to enable ML engineers to sufficiently conduct model development and experiments. Additionally, we build and operate our own data pipeline including data collection and preprocessing using cloud services. We also work with various software engineers (backend/frontend/DevOps/MLSE) who help productionize ML models.
Responsibilities
ML Engineers at Match Group AI are scientists who research and apply cutting-edge AI/ML technology, and engineers who design and optimize models and systems to fit actual service environments. The organization handles various tasks as follows, and we are looking for someone who can skillfully perform at least one of these:
- Solve various business problems that arise in products serviced by Match Group. We understand the background, goals, and constraints of business problems, redefine them as correct AI/ML problems, and execute the process of solving them by finding the best methodology.
- Participate in the development of new products and features at Match Group. From ideation through prototyping to reaching actual users, we execute the process of implementing products and features quickly and correctly using AI/ML technologies.
- Lead research and development of large language models and multimodal models. We create models specialized in the dating domain using Match Group's data and cutting-edge technologies, and explore ways to apply them across various areas of the product.
To perform the above tasks, Sr. ML Engineers are expected to have solid AI/ML foundational knowledge and the ability to lead the team's technical decision-making based on rapid acquisition of the latest technologies. Based on these capabilities, you will primarily lead the following technical topics:
- Methods to correctly handle data with various modalities including text, images, and event logs, and methods to correctly resolve bias and noise in collected data
- Definition of evaluation metrics aligned with product goals and optimal model training methods to achieve goals
- On-device and large-scale model inference optimization methods considering engineering constraints and infrastructure environments
Requirements
- Possess understanding of the AI/ML domain in general and in-depth knowledge of at least one specific domain, with a PhD in a related field or 5+ years of industrial project experience, or equivalent capabilities
- Able to define ambiguous business or technical problems as correct AI/ML problems and solve them based on sound technical decision-making
- Able to quickly perform development using open-source frameworks such as PyTorch, Tensorflow, JAX, and general Python development, and manage high-quality code for collaboration and maintenance
- Able to design stable ML training pipelines and deployment processes
- Able to continuously explore new technologies, share them, and help the team maintain high-level AI/ML technical competitiveness
- Able to lead technical decision-making in AI projects involving two or more ML Engineers
- Possess excellent communication skills to collaborate with stakeholders from various departments
- Able to communicate smoothly in Korean and capable of basic communication in English
Preferred Qualifications
- Have publications in top-tier AI/ML conferences and journals (NeurIPS, ICLR, ICML, CVPR, ICCV/ECCV, KDD, โฆ) or award-winning achievements in related competitions
- Have experience leading the entire process required when training LLMs or similarly-scaled models at large scale and creating results
- Possess large-scale software development experience and engineering capabilities necessary for architecture/system design
- Able to plan A/B test experiments, correctly define target KPI metrics, and professionally perform SQL-based data analysis
- Able to discover meaningful insights from data using causal analysis and other statistical techniques and correctly apply them to decision-making
- Have experience leading engineering teams
- Have experience as a PO/PM or equivalent
- Able to conduct real-time business communication fluently in English
Hiring Process
Employment Type: Full-time Hiring Process: Document Review > Pre-assignment > Live Coding Interview > Technical Interview > Technical In-depth and Cultural Alignment Interview > Final Offer (* The hiring process may be modified if necessary.) For the document review stage, successful candidates will be notified individually. Application Materials: Detailed career-based English resume in free format (PDF)
This position is eligible for hiring of professional researchers on active duty/career change, professional researchers on supplementary service/career change. For military service exception personnel, service management will be conducted in accordance with relevant military service exception laws.
If there are false statements in the submitted materials or if there are grounds for disqualification from providing labor under relevant laws, the hiring may be cancelled. If necessary, additional screening and document verification may be conducted beyond the hiring process previously announced. Veterans are given preferential treatment in accordance with relevant laws. If you are eligible, please notify us when applying and submit supporting documents at the time of hiring.
When applying for positions hired by Hyperconnect, the following privacy policy applies to personal information processing: https://career.hyperconnect.com/privacy
#HPCNT