Sr. Machine Learning Software Engineer

Key Skills

Aartificial intelligence, Machine learning techniques, Optimization

Job Description

ML Software Engineer aims to apply all of our AI technologies to products to create business impact and develop sustainable systems to accelerate the application of AI technologies. To achieve this goal, we (1) develop scalable backend servers based on ML models, (2) develop/operate real-time data pipelines for ML model inference, and (3) collaborate with other teams to platformize ML components that can be managed in common. In the process, we work closely with other departments and proactively participate in all processes (problem definition, hypothesis setting, experimental design, analysis, and feedback) required to achieve KPIs. For more detailed information about our work, please refer to the following content.

• AI Platform Dev Team Interview
• ML Application Team Interview
• Head of AI - Shurain Interview
• AI in Social Discovery (Blending Research and Production)

Interesting problems that ML Software Engineers are solving are uploaded to the Tech blog .

• Compute real-time ML feature data using ksqlDB
• 10 Python Performance Optimization Tips for High-Performance ML Backends
• Reduce machine learning model serving costs by 1/4
• Operate an event-based live streaming recommendation system
• Data storage technology for machine learning applications

Work environment

[Domestic top-level AI organization]

We work with Machine Learning Engineers and Machine Learning Research Scientists who regularly publish papers in top-tier AI/ML conferences. You can check out the papers published by Hyperconnect here.

[Rich MLOps Know-How]

You will work in an organization with a wealth of MLOps know-how, with over 50 models in production.

• Deview 2023 - Feature Store Implementation for Real-Time Recommendation Systems
• Deview 2021 - How to deploy more models faster?
• Building a high-performance deep learning cluster, Part 1

Responsibilities

Hyperconnect is working in various ways to apply machine learning technology to its products. Hyperconnect’s ML Software Engineers will perform one or more of the following four major tasks:

[Machine Learning-Based Backend Application Design and Implementation]

We develop various machine learning-based backend services (FastAPI, Spring) to improve the quality of services operated by Hyperconnect and Match Group. They are designed with a lot of consideration from a performance perspective to enable real-time operation on a global scale, and the microservices operated by the team are handling the highest level of traffic within the company.

• Personalized recommendation system and search system (Azar, Hakuna, services within Match Group)
• Development of AI Flywheel system that continuously improves models
• Development of a moderation system and platform for Trust & Safety
• In-house API server that converts unstructured data such as video, voice, and natural language into an interpretable form

[Developing a real-time data pipeline for model inference]

We develop pipelines (Apache Flink, KSQL) that process real-time events and use them for model inference. We consider and design systems (ex. streaming applications, feature stores) to quickly and reliably collect, process, and serve features. Sometimes, we proactively discover features that improve model performance during the pipeline development process.

[Development of client SDK with machine learning capabilities]

We develop and operate machine learning SDKs that are embedded in Hyperconnect and Match Group's mobile applications. The SDKs developed by Hyperconnect must be able to operate on numerous devices used around the world, so they are being developed with a very high target of stability. In addition, they must be able to be continuously updated to provide the highest-performance machine learning API, and for this purpose, we are also developing a scalable backend platform.

[Machine Learning-Based New Service PoC]

Hyperconnect is researching and developing various machine learning technologies that contribute to products. In order to measure the business impact of new machine learning models developed in-house, Proof of Concept (PoC) is often required, and MLSE also plays a role in developing PoC products. By developing PoC products, the development cycle is shortened and more iterations are performed, allowing model improvements to be made faster.