Founding Staff AI Engineer — Physics AI, Agentic Systems & Full-Stack

About Elucida Labs Elucida Labs is building the autonomous AI Scientist for Semiconductor Manufacturing.Advanced wafer fabrication, spanning thin-film deposition (ALD, CVD, PVD), plasma etch (RIE, ALE), CMP, and advanced packaging, is bottlenecked by brutal multi-month empirical trial-and-error cycles. We build intelligent software that eliminates months of guesswork and trial-and-error in semiconductor manufacturing.Backed by top founders and mentors (Stanford StartX F26), we work with semiconductor equipment OEMs, microelectronics research institutes, and Fabs. As a Founding Staff AI Engineer, you will own the technical architecture from core scientific models to live customer deployments.The Role We are seeking an exceptional, high-agency Founding Staff AI Engineer who thrives at the intersection of deep scientific modeling, autonomous agentic architectures, and rapid production full-stack engineering.We operate with extreme engineering leverage: our team heavily uses agentic developer tooling, including Claude Code, Cursor, and the Model Context Protocol (MCP), to ship robust, production-grade systems at 10x startup velocity. You will bridge physical process models and large-scale autonomous LLM agents, write full-stack code, and deploy directly into customer evaluations and pilots.Key Responsibilities 1. Physics AI & Scientific Machine LearningDesign and implement Physics-Informed Neural Networks (PINNs), neural operators (FNOs, DeepONets), and surrogate models to simulate plasma dynamics, chemical kinetics, thermal budgets, and material interfaces.Build active learning and Bayesian optimization pipelines to recommend optimal recipe parameters (gas flow, RF power, pressure, temperature profiles) under sparse, high-cost experimental constraints.Ingest and featurize complex, multimodal semiconductor fab telemetry: optical emission spectroscopy (OES), residual gas analysis (RGA), in-situ metrology, and wafer-level morphology data. 2. Agentic AI & Autonomous Scientific WorkflowsDevelop robust, deterministic multi-agent systems that autonomously decompose complex engineering goals into actionable experimental design, data querying, and simulation workflows.Build tool-use capabilities and MCP (Model Context Protocol) servers allowing agents to execute scientific code, run numerical solvers, and synthesize technical papers/patents.Implement structured reasoning, guardrails, state-machine verifiers, and memory stores to ensure agent outputs meet rigorous physical plausibility and safety criteria. 3. Full-Stack Engineering & AI-Augmented VelocityArchitect, build, and maintain our end-to-end software platform: high-performance backend microservices (Python/FastAPI/AsyncIO), database layers, and intuitive engineering dashboards (TypeScript/React/Next.js).Leverage agentic coding tools (such as Claude Code CLI and modern agent environments) to autonomously write test suites, refactor codebases, automate CI/CD tasks, and maximize raw development output.Construct robust ETL pipelines and secure data connectors capable of ingesting high-frequency equipment time-series and legacy SECS/GEM or flat-file fab logs. 4. Customer Delivery & Pilot ExecutionOwn technical deployment and integration during pilot programs with equipment OEMs and cleanroom customer sites.Work shoulder-to-shoulder with customer process engineers, CTOs, and fab managers to translate complex domain challenges into working product capabilities.Deploy air-gapped or hybrid on-prem/VPC customer instances ensuring enterprise-grade data isolation, encryption, and zero-data-leakage compliance. QualificationsScientific & Applied ML Background: Proven experience building and deploying scientific ML models, PINNs, surrogate models, or complex physical simulation workflows (fluids, plasma, materials, or chemical engineering domains).Agentic & LLM Expertise: Hands-on experience building production-grade agentic frameworks (LangGraph, AutoGen, CrewAI, or bespoke state machines), advanced tool-calling, evaluation frameworks, and retrieval-augmented generation (RAG).AI-First Developer Leverage: Daily fluency with agentic coding environments (especially Claude Code, Cursor, or custom terminal agents) and protocols like MCP to accelerate prototyping, multi-file refactors, and test coverage.Full-Stack Competence: Proficiency with modern backend and frontend stacks (Python/FastAPI, TypeScript/React/Next.js, Node.js, Tailwind, GraphQL/REST).Data & Systems Infrastructure: Experience with relational and time-series databases (PostgreSQL, TimescaleDB, ClickHouse), vector databases, Docker, Kubernetes, and cloud infrastructure (AWS/GCP/Azure).SaaS product deployments: Hands-on experience managing SaaS product deployments, CI/CD pipelines, and cloud infrastructure (AWS/GCP/Azure)Startup Velocity & Customer Focus: Ability to navigate ambiguity, ship working code daily, iterate directly based on customer feedback, and take extreme end-to-end ownership.U.S. Person: Due to US export control regulations, this position strictly requires the applicant to be a 'U.S. Person' (a U.S. Citizen, U.S. Permanent Resident, or approved Asylee/Refugee). Bonus Points· Advanced Degree: Ph.D. from a Tier-1 university (e.g., Stanford, MIT, UC Berkeley, CMU, Caltech, UIUC, or top international equivalent) in Computer Science, Applied Physics, Materials Science, Chemical Engineering, Electrical Engineering, or related technical disciplines.· Core Compute/Systems: Experience with high-performance computing and physics engines (C++).· AI & Physics Modeling: Familiarity with physics-informed machine learning and AI frameworks (NVIDIA Modulus, NVIDIA NeMo, PyTorch).· Semiconductor Domain Knowledge: Prior background or academic research in semiconductor processing, nanotechnology, thin films (ALD, CVD, Etch, CMP), or electronic materials.· Fab Protocols & Tool Interfacing: Experience interfacing with hardware control systems, industrial IoT, or fab communication protocols (SECS/GEM, EDA/Interface A).· Tooling & Protocol Extensions: Experience authoring custom MCP servers or building integrations for agentic coding tools.· Early-Stage Experience: Previous experience as a founder, founding engineer, or early-stage engineer in a venture-backed deep-tech or enterprise AI startup. Why Join Elucida Labs?Pivotal Ownership: Ground-floor equity and outsized influence in setting company architecture, culture, and technical vision.Massive Industrial Impact: Directly accelerate the physical limits of semiconductor fabrication, advanced packaging, and microelectronics manufacturing.World-Class Ecosystem: Tight access to top-tier semiconductor industry executives, fab leaders, and the Stanford StartX network.