AI Software Engineer (C++)

📋 Description

• • G2i Inc. is seeking a highly skilled and detail-oriented AI Software Engineer with a strong C++ background to join our innovative team. This fully remote role focuses on enhancing the capabilities of large-language models (LLMs) specifically for generating production-grade code across a diverse array of programming languages. You will play a pivotal role in the Reinforcement Learning from Human Feedback (RLHF) pipeline, directly contributing to the development of AI that can propose, critique, and improve code with the same rigor and insight as experienced human engineers.
• • Your primary responsibility will be to meticulously evaluate and refine AI-generated code. This involves a critical assessment of code snippets, where you will compare and rank multiple solutions, providing clear, concise, and well-reasoned explanations for your choices. This deep analysis will help the AI understand the nuances of good coding practices, efficiency, and correctness.
• • A significant part of your work will involve repairing and refactoring code produced by the AI. You will identify and correct errors, optimize for performance, and ensure adherence to established coding styles and best practices. This hands-on approach to code improvement is crucial for training the model to produce high-quality, shippable code.
• • You will be instrumental in injecting expert feedback into the RLHF pipeline. This includes providing ratings, making edits, and analyzing test results. Your ability to articulate the 'why' behind your feedback will be paramount, as this data is converted into reward signals that guide the reinforcement learning process.
• • The RLHF process, in essence, involves generating code, having expert engineers like yourself rank, edit, and justify their feedback, converting this valuable input into quantifiable reward signals, and then using reinforcement learning to tune the model. The ultimate goal is to create an AI that generates code you would confidently ship to production.
• • This role demands an exceptional level of attention to detail. You must be adept at spotting subtle logic errors, potential performance bottlenecks, and security vulnerabilities. Your ability to perform thorough code reviews will be directly leveraged to train the AI to avoid these pitfalls.
• • Excellent written communication skills are non-negotiable. You will frequently need to articulate complex technical reasoning, explaining why one code solution is superior to another. This clarity in communication is fundamental to the effectiveness of the training process.
• • You should possess a strong aptitude for understanding technical documentation and language specifications. This role thrives in an asynchronous, low-oversight environment, requiring self-motivation and the ability to work independently while delivering high-quality results.
• • While no prior experience with RLHF or AI training is necessary, a solid foundation in software engineering principles is essential. We are looking for engineers who can apply their existing code review instincts and analytical skills to the AI training domain. The company will provide the necessary training on the specifics of AI model training and RLHF.
• • The core technology stack for this role is C++. A deep understanding and practical experience with C++ are critical for effectively evaluating and refining code across various programming languages.
• • This is a unique opportunity to contribute to the cutting edge of AI development in software engineering. You will be part of a team that is shaping the future of how code is generated and improved, making a direct and tangible impact on the capabilities of advanced AI models.
• • The role is structured as a 1099 independent contractor position, offering flexibility in terms of hours, with a minimum commitment of 15 hours per week and up to 40 hours available. Compensation is competitive, ranging from $30 to $70 per hour, dependent on your location and level of experience.
• • You will be working in a fully remote capacity, allowing you to work from anywhere within the accepted countries and locations listed in the provided document. This offers unparalleled flexibility and work-life balance.
• • The company values straightforward impact and minimal bureaucracy. If you are a seasoned C++ engineer with a passion for code quality and a keen eye for detail, and you are excited by the prospect of training AI to write better code, this role presents a compelling opportunity to apply your expertise in a groundbreaking field.