Staff Software Engineer (Forward Deployed) at Invisible Technologies in New York, NY

Job description

About Invisible Invisible is an Applied AI research company closing the gap between frontier models and the enterprise. We provide the platform infrastructure that makes AI work in production at scale: unify fragmented data, map complex processes, orchestrate and evaluate agents, and stand up hyper-personalized workflow applications fast for companies like Grubhub, Alliance Global, and Thomson Reuters. Where models fall short, we build RL environments and data to fix them. The result: a virtuous cycle between enterprise deployment and model training. Recognized as Best Predictive Analytics Platform of the Year at the 2026 AI Breakthrough Awards, Invisible makes AI work. About The Role As a Staff Forward Deployed Engineer (FDE) at Invisible, you'll lead high-impact, AI-powered solutions that reshape how our clients operate their most critical workflows. You won't just build and deploy — you'll drive the strategy, architecture, and execution of end-to-end systems, working directly with clients. This is a hybrid role: equal parts AI engineer, software builder, and technical consultant. It's perfect for someone who wants to be hands-on with models and close to the impact they generate. The team is increasingly focused on applied AI work — building and deploying LLM-powered and agentic systems in production — so strong ML engineering instincts can be a valuable addition alongside your architecture and leadership experience. What You’ll Do Partner with delivery and executive stakeholders to scope, design, and lead implementation of AI-driven solutions, including LLM and agentic systems Identify transformational opportunities in messy, ambiguous workflows and turn them into repeatable systems Lead architecture design and trade-off discussions across performance, scalability, cost, and reliability — including for ML/AI-specific systems like RAG pipelines and production-grade AI applications Own projects from first discovery call through full deployment — including client-facing delivery, internal coordination, and post-launch iteration Build shared infrastructure, reusable components, and internal playbooks to level-up the team, including MLOps practices for model versioning, monitoring, and lifecycle management Coach and mentor mid-level engineers and help shape the culture of forward-deployed AI engineering at Invisible What We Need 8+ years of software engineering experience, including significant time spent building data, ML, or backend systems Python & ML/LLM Frameworks: Deep proficiency in Python with hands-on experience using Hugging Face, LangChain, OpenAI, Pinecone, and related ecosystems, including LLMs and AI Agents Deployment & Infrastructure: Skilled in full-stack and API-based deployment patterns, including Docker, FastAPI, Kubernetes, and cloud environments (GCP, AWS, Azure) Platform Orchestration: Experienced with workflow orchestration libraries, pub/sub systems (Kafka), and schema governance Data Management: Expertise in data governance and operations, including Unity Catalog and policy management, cluster/job orchestration, data contracts and quality enforcement, Delta/ETL pipelines, and replay processes ML Operations (for teams with an applied AI/ML focus): Experience with MLOps frameworks and best practices — model versioning, monitoring, and lifecycle management — and designing scalable ML systems such as RAG pipelines Strong product and system design instincts — you understand business needs and how to translate them into technical architecture Experience building usable systems from messy data and ambiguous requirements Excellent communication and client-facing skills; you've led conversations with technical and non-technical stakeholders alike Proven experience owning projects from scoping through deployment in ambiguous, high-stakes environments Be willing to be on-call for our customers when situations arise Travel: Willingness to travel for on-site client engagements as needed Onsite: Available to work from