Note: The job is a remote job and is open to candidates in USA. Element 84 is a woman-owned small business that develops geospatial data processing pipelines and builds software for various sectors. The AI Systems Engineer will design and build multi-agent systems for NASA's Text-to-Spaceship initiative, ensuring the generated spacecraft components are safe and manufacturable.
Responsibilities
- Physics-Informed AI Constraints: Translate physical constraints (mass, thermal envelopes, radiation shielding, vibro-acoustics) and mechanical properties into prompt contexts, structured outputs, and agent guardrails
- Automated CAE & FEA Workflows: Design and implement programmatic workflows for Finite Element Analysis (FEA) and computational fluid dynamics (CFD) that the AI agents can autonomously trigger to validate their own designs
- Domain Validation & Ground Truth: Serve as the human-in-the-loop expert to vet AI-generated structural designs, identifying edge cases, hallucinated physics, or impractical geometries, and using those insights to improve the agentic evaluation framework
- Multi-Agent Architecture: Design and implement agentic frameworks where a lead orchestrator agent coordinates specialized agents (optical design, structural, harnessing, analysis, reporting) across complex, multi-step engineering workflows
- Natural Language to Engineering Output: Build pipelines that convert natural language mission requirements into structured specifications (text → JSON) and implement RAG pipelines for engineering knowledge retrieval
- Tool Integration: Connect AI agents to external engineering tools (CAD, FEA, CAM software) via MCP and custom API integrations, enabling agents to drive design, analysis, and manufacturing workflows
- Cloud Infrastructure: Deploy and scale AI workloads across cloud providers (AWS, GCP, Azure) using containerized architectures. Apply cloud security best practices for government data
- Evaluation & Observability: Build evaluation frameworks for agentic applications — measuring agent performance, design quality, and pipeline reliability across multi-step autonomous workflows
Skills
- Aerospace or Manufacturing Experience: 2+ years working as a software engineer within the aerospace, defense, space, or similar manufacturing sectors, with a strong understanding of the hardware engineering lifecycle and launch and spaceflight environment
- Hands-on experience with the scripting APIs of industry-standard engineering tools (e.g., Python APIs for Autodesk Fusion 360, ANSYS, NASTRAN, or SolidWorks)
- Familiarity with aerospace-grade materials (titanium, aluminum alloys, carbon fiber composites) and how their properties dictate design limits
- Ability to interpret and programmatically apply GD&T (Geometric Dimensioning and Tolerancing) and NASA/aerospace engineering standards
- Experience with AWS services and architecture. Familiarity with other cloud providers (GCP, Azure) is a plus
- Real-world experience with Python. Secondary proficiency in TypeScript is a plus
- Strong grasp of API design, containerization, and connecting heterogeneous tools and data formats into automated pipelines
- Experience with IaC tools and reproducible cloud deployments
- Deep understanding of how large language models work — context windows, structured output, prompt engineering, and model selection trade-offs
- Ability to design and implement evaluation strategies for complex LLM workflows, measuring correctness, reliability, and performance of multi-step autonomous systems
- Translate physical constraints (mass, thermal envelopes, radiation shielding, vibro-acoustics) and mechanical properties into prompt contexts, structured outputs, and agent guardrails
- Design and implement programmatic workflows for Finite Element Analysis (FEA) and computational fluid dynamics (CFD) that the AI agents can autonomously trigger to validate their own designs
- Serve as the human-in-the-loop expert to vet AI-generated structural designs, identifying edge cases, hallucinated physics, or impractical geometries, and using those insights to improve the agentic evaluation framework
- Design and implement agentic frameworks where a lead orchestrator agent coordinates specialized agents (optical design, structural, harnessing, analysis, reporting) across complex, multi-step engineering workflows
- Build pipelines that convert natural language mission requirements into structured specifications (text → JSON) and implement RAG pipelines for engineering knowledge retrieval
- Connect AI agents to external engineering tools (CAD, FEA, CAM software) via MCP and custom API integrations, enabling agents to drive design, analysis, and manufacturing workflows
- Deploy and scale AI workloads across cloud providers (AWS, GCP, Azure) using containerized architectures. Apply cloud security best practices for government data
- Build evaluation frameworks for agentic applications — measuring agent performance, design quality, and pipeline reliability across multi-step autonomous workflows
- Experience building agents with agentic libraries like Pydantic AI
- Expertise in the Model Context Protocol (MCP) or equivalent approaches for connecting AI agents to external APIs, databases, and domain-specific tools
- Experience with ML techniques beyond LLMs (e.g., reinforcement learning, uncertainty quantification, reduced-order models) applied to design optimization or engineering problems
- Contributions to open-source AI libraries or a portfolio of deployed LLM applications
Benefits
- Competitive medical, dental and vision benefits
- Life Insurance, Short & Long Term disability insurance
- Voluntary Accident, Critical Illness & Hospital Insurance
- 401(k) and Roth 401(k) retirement plans with a fixed 3% of salary employer contributions (paid regardless of employee participation)
- Health savings account with a company contribution
- Flexible spending accounts (medical, dependent care and transportation)
- Company-paid parental leave after one year of employment
- Flexible work schedules
- Paid employee assistance program
- 6 paid floating holidays
- 4 weeks + 1 day paid Vacation Time Off per calendar year (prorated first year)
- 40 hours paid Sick Leave
- Cell phone stipend
Company Overview