Software Engineer, AI Native

Meta•Menlo Park, CA

1d•$59 - $181,000

About The Position

Meta is seeking talented engineers to join our teams in building cutting-edge products that connect billions of people around the world. As an AI Native SWE, you will work on complex technical problems, build new AI-powered and generative AI features, and improve existing products across all platforms. Our teams are pushing the boundaries of user experience through LLMs, conversational and multi-modal AI, context-aware systems, and AI-powered automation—and we’re looking for engineers who bring an AI-first mindset, move fast through rapid iteration and experimentation, and raise the bar on quality and reliability for AI-driven experiences.

Requirements

Currently has, or is in the process of obtaining a Bachelor's degree in Computer Science, Computer Engineering, relevant technical field, or equivalent practical experience. Degree must be completed prior to joining Meta
2+ years of programming experience in a relevant language OR a PhD + 9 months programming experience in a relevant language
Experience building maintainable and testable codebases, including API design and unit testing techniques
Experience effectively utilizing AI technologies and tools (e.g., large language models, agents, etc.) to enhance workflows

Nice To Haves

Experience with one or more languages such as C/C++, Java, Python, JavaScript, Hack, and/or shell scripting
Experience improving quality through thoughtful code reviews, appropriate testing, rollout, monitoring, and proactive changes
Experience with architectural patterns of large-scale software applications and improving efficiency, scalability, and stability of system resources
Experience with ML tooling/frameworks such as PyTorch, TensorFlow, and Python
Experience with AI/ML techniques and workflows such as fine-tuning, transfer learning, few-shot/zero-shot approaches, and/or model distillation
Experience in one or more of the following: LLMs, generative AI, machine learning, recommendation systems, pattern recognition, data mining, or related fields
Experience implementing RAG, embeddings, or knowledge-backed generation and familiarity with tokenization and transformer-based systems
Experience designing AI agents, orchestration, and human-in-the-loop systems and treating AI as a collaborator to accelerate delivery
Understanding of Responsible AI practices (AI safety, ethics, alignment, explainability) and building safeguards/quality controls for AI outputs
Experience building and utilizing AI technologies (e.g., LLMs, agents, orchestration systems) as collaborative tools to streamline workflows and accelerate delivery

Responsibilities

Collaborate with cross-functional teams (product, design, operations, infrastructure) to build innovative AI-native application experiences
Build and integrate LLM / generative AI capabilities into product surfaces (mobile, web), including prompt engineering, structured prompting, and context management
Develop and maintain reusable software components for interfacing with back-end platforms, model serving/inference layers, and AI toolchains
Implement retrieval-augmented generation (RAG) patterns (e.g., embeddings + retrieval) and contribute to context-aware and personalized user experiences
Design/Contribute to agentic workflows and leverage AI tools and agents (including human-in-the-loop / expert-in-the-loop designs) to automate tasks and scale impact
Analyze, debug, and optimize code and systems for quality, efficiency, performance, reliability, and cost
Establish effective quality practices for AI features, including evaluation/QA for AI outputs, monitoring, and iterative improvement via feedback loops
Architect efficient and scalable systems that power complex applications and AI-enabled features, identify and resolve performance and scalability issues
Drive end-to-end execution of medium-to-large features with increasing independence, contribute to technical direction within the team
Establish ownership of components, features, or systems with comprehensive end-to-end understanding