Senior Generative AI Engineer | 100% Remote (Global) The Hook: Build the Brains of the Future Workforce At the intersection of logic and creativity, we aren't just building a tool; we are building the next generation of "colleagues." We believe that the future of software engineering isn't just assisted by AI—it is driven by it. As a Senior Generative AI Engineer, you won’t be polishing chatbots. You will be the architect of autonomous coding agents capable of reasoning through complex PRs, refactoring legacy debt, and ship-ready code. If you’ve ever felt that current LLMs are just "scratching the surface" of what’s possible in software automation, this is where you prove it. Role Overview As a Series A startup, we operate on the bleeding edge. We embrace radical transparency—meaning every decision, success, and "fast break" is shared openly across our global team. We move at a pace that would make most enterprise companies dizzy, and we need a Senior Engineer who treats ambiguity as a playground rather than a hurdle. You will report directly to the CTO and own the end-to-end lifecycle of our agentic reasoning engines. What You Will Do We don’t believe in siloed engineering. Here is how you will apply your expertise to our mission: Architecting Agentic Reasoning: You won’t just "call an API." You will be utilizing Python and PyTorch to develop custom training loops and inference strategies that allow our agents to "think" before they code. Fine-Tuning for Precision: When off-the-shelf models fail at complex syntax, you’ll step in to lead fine-tuning efforts on LLaMA and Mistral. You’ll be responsible for curating high-quality synthetic datasets to improve our agents' ability to handle niche languages and frameworks. Scaling Context with RAG: You will design and optimize high-performance vector retrieval systems using Pinecone or Milvus. Your goal is to ensure our agents have "perfect memory" of massive, multi-million line codebases, providing them with the context needed to make surgical-grade edits. Rapid Prototyping: In the spirit of "moving fast and breaking things," you will ship experimental features to production weekly. You'll evaluate model performance in real-world scenarios, discarding what fails and doubling down on what works. Requirements Technical Mastery: Deep proficiency in Python and PyTorch. You should be comfortable navigating the internals of transformer architectures. LLM Specialization: Proven experience in fine-tuning (SFT, RLHF, or DPO) open-source models like LLaMA and Mistral to achieve specific, high-accuracy benchmarks. Vector Infrastructure: Experience building and scaling RAG pipelines with Pinecone, Milvus, or Weaviate, specifically focusing on latency and retrieval precision. High Autonomy: You are a "company of one." You can take a vague objective ("Make the agent better at debugging") and turn it into a technical roadmap without hand-holding. Research-to-Production Mindset: You stay updated on the latest ArXiv papers, but your primary joy comes from seeing those theories improve a live product. Why Join Us True Global Autonomy: We are 100% Remote. We don’t care where you are or what your "office" looks like; we care about the quality of your code and the speed of your iterations. Our culture of radical transparency means you have access to the same information as the founders. The "Move Fast" Philosophy: We value speed over perfection. Here, you have the permission to take risks, break things, and learn at a 10x rate compared to traditional tech firms. Equity & Impact: As an early hire in a fast-growing Series A, you’ll have a significant equity stake. You aren’t just a number; you are a cornerstone of the company. Collaboration without Borders: Despite being remote, we foster a tight-knit "async-first" culture. We prioritize deep work and documentation, ensuring you spend less time in meetings and more time building.