AI Agent Engineer - TikTok LIVE (Singapore)

We value your passion for technology, understanding and insight into business, as well as past achievements. We hope you possess excellent fundamental capabilities and growth potential:

• Strong engineering and implementation skills

Possess excellent coding ability and solid software engineering fundamentals. Familiar with at least one mainstream backend development language (e.g., Go, Python, Java, Rust, C++), and capable of independently designing and implementing complex systems.

• Familiarity with the LLM and Agent technology ecosystem

Have a deep understanding of large language models (LLMs), be familiar with mainstream Agent paradigms, and have practical experience integrating AI technologies into real-world workflows (partial experience is also acceptable).

• Excellent problem decomposition and system design skills

Able to clearly break down and abstract complex, ambiguous business problems into scalable system architectures and reusable technical solutions.

• Strong technical curiosity and learning ability

Maintain a high level of enthusiasm for cutting-edge technologies, continuously follow industry best practices, and quickly learn and apply new knowledge to practical work.

• Hands-on project experience with AI Agents or related fields, including leading or deeply participating in Agent design and implementation.
• In-depth understanding or work experience in performance analysis, observability, or data analysis.
• Basic knowledge of the live streaming technology stack (e.g., audio/video, network transmission, rendering).
• Active participation in high-quality open-source projects or recognized influence in the technical community.

• Work at the forefront of technology: Help build industry-leading performance analysis agents and redefine the paradigm of performance optimization in the live streaming domain.
• Create real business impact: Your work will be applied to one of the world’s largest live streaming platforms, solving complex, high-impact technical challenges.
• Unlock significant growth opportunities: Grow in a fast-moving, innovation-driven global team, and collaborate with top talent worldwide.

We believe the best talent is driven to solve the most important problems using cutting-edge technologies. If you are passionate about AI, performance optimization, and delivering exceptional user experiences, we look forward to building a smoother and smarter LIVE future with you. Core Responsibilities As a core member of the team, you will be deeply involved in the design, R&D, and platform implementation of the Performance Analysis Agent, transforming cutting-edge AI technologies into a powerful engine that drives business growth.

• Design and develop the core capabilities of the Performance Analysis Agent

Lead or contribute to developing AI-powered capabilities for the performance analysis workflow, including but not limited to:

• Automatic Data Inspection and Attribution: Real-time monitoring, anomaly detection, and minute-level automated attribution of massive performance metrics.
• Automation of Experimental Analysis: Automatically generate scientific and rigorous A/B experiment reports to improve decision-making efficiency and quality.
• Thematic Analysis Assistance: Provide AI support for complex performance topics (e.g., power consumption, smoothness, loading speed) and produce in-depth insights at scale.
• Shift-Left Code Performance Review: Integrate AI into the Code Review (CR) phase to automatically detect performance risks, identify issues early, and address them at the source.
• Build and evolve the Agent platform and services

Contribute to the development and evolution of the team’s Agent platform, upgrading traditional APIs to a more powerful and flexible MCP (Model-Controller-Proxy) service framework. Strengthen the data and tooling foundation to enhance overall scalability, reusability, and reliability.

• Continuously improve Agent core performance

Leverage industry-leading technologies to enhance the Agent’s accuracy and efficiency in understanding, retrieval, decision-making, and response, including:

• Prompt Engineering and Chain of Thought (CoT): Design and optimize prompts to fully unleash large model capabilities.
• Retrieval-Augmented Generation (RAG): Combine with high-quality knowledge bases to improve accuracy and timeliness of responses.
• Function Calling: Extend the Agent’s ability to interact with external tools and data sources.
• Memory Mechanism and Evaluation Iteration: Enable the Agent to learn and iterate while establishing a robust, scientific evaluation system.
• Drive the AI value closed loop

Deeply understand live streaming business and technical pain points, identify opportunities for AI-driven innovation, and work closely with business, product, and R&D teams to turn AI solutions into measurable outcomes—improving R&D efficiency, service quality, and user experience metrics.