Position Overview We are looking for a highly motivated Software Engineer to join the LLMBoost team at MangoBoost. LLMBoost is our enterprise-grade platform for large language model deployment and optimization, supporting inference, training, fine-tuning, and retrieval-augmented generation across both AMD and NVIDIA GPUs. You will work alongside a team of experts to build, optimize, and scale AI workloads for high-performance, multi-node environments.

Responsibilities & Opportunities

• Develop, optimize, and test features for the LLMBoost platform, including model deployment, auto-tuning, and multi-GPU orchestration
• Contribute to performance benchmarking and optimization on both AMD and NVIDIA hardware
• Maintain infrastructure for automated builds, testing, and releases
• Collaborate with the team on debugging, profiling, and improving system reliability

Required Qualifications

• BS/MS/PhD in Computer Science, Electrical Engineering, or a related field (or equivalent experience)
• Proficiency in Python and Linux-based development environments
• Strong problem-solving skills and eagerness to work on large-scale AI systems

Preferred Qualifications

• Can squeeze extra performance out of GPUs through kernel tuning and advanced profiling techniques
• Have published research or developed innovations in inference optimization, training acceleration, or distributed AI systems
• Know your way around Kubernetes and Docker for managing large-scale, containerized AI workloads
• Have hands-on experience running multi-node GPU training and inference jobs
• Understand the ins and outs of MLPerf or other AI benchmarking suites
• Can spot and fix network or I/O bottlenecks in high-throughput AI pipelines
• Have contributed to open-source AI/ML frameworks and love improving the tools the community relies on

About Us MangoBoost, Inc.

• MangoBoost is a start-up company that develops groundbreaking new data processing units (DPUs). MangoBoost’s novel DPUs can be customized and flexibly integrated in server systems to address a myriad of system overheads, and make other devices (such as SSDs, GPUs, NICs, etc.) smarter. As a result, MangoBoost DPU offers a dramatic boost in system efficiency, performance, scalability, and cost effectiveness. MangoBoost leverages ~10 years of prior world-leading research done in academia, which has produced extensive full-system patent-pending prototypes published at premier conferences in computer architecture and systems areas (for example ISCA and OSDI). MangoBoost raised a $10M seed round fund in 2022 and a $55M series A round in 2023 and is actively growing in the US and Korea.
• MangoBoost is comprised of the world's top computer system developers. The founder is currently one of the world's leading computer system architects, and other members are also experts with outstanding careers in various aspects of computer system development. Most team members have conducted research and worked at top-tier universities and companies worldwide. Since 2014, our team has been developing the world's top-level DPU, and we are the only start-up capable of designing a DPU that can compete immediately with giants like Amazon and NVIDIA.

Notice Regarding Return of Hiring Documents

• This notice is in accordance with Article 11, Section 6 of the Fair Hiring Procedure Act, which allows job applicants, excluding the final successful candidate, to request the return of any recruitment documents they have submitted.
• Job applicants who were not selected as the final candidates in the recruitment process are entitled to request the return of the recruitment documents they submitted within 180 days of the date the employment decision is confirmed. However, this does not apply to documents submitted via the website or email, or documents voluntarily submitted by the applicant without the company’s request. If the recruitment documents are lost due to natural disasters or other reasons for which the company is not responsible, it will be considered as having been returned.