Role Overview
We are seeking an experienced AI Architect to design, lead, and deliver an AIdriven 2D/3D tolerance stackup analysis platform for aerospace and complex mechanical systems.
The role focuses on engineeringgrade accuracy, agentic AI architectures, and fully onpremise deployments with no public cloud dependency.
The architect will own the endtoend AI and system architecture for stackup analysis, integrating CAD systems, engineering drawings, mathematical models, and GenAIbased agents, while ensuring scalability for future engineering use cases.
Key Responsibilities
Architecture & Solution Design
- Define and own the overall AI and system architecture for stackup analysis using an agentic, microservicesbased platform
- Design onpremise, isolated deployments compliant with enterprise IT, security, and dataprivacy policies
- Create a scalable and futureready architecture that can onboard additional engineering agents without core redesign
StackUp Analysis AI Capabilities
- Architect AI solutions for 2D and 3D tolerance stackup analysis, including:
- Axial and radial clearance calculations
- Worstcase and RSS methods
- Multijoint and multipart tolerance aggregation
- Enable support for CATIA V6, Autodesk CAD, and PDF engineering drawings
- Design mechanisms to:
- Handle varied dimensioning schemes
- Flag lowconfidence results
- Suggest manufacturabilitydriven tolerance improvements
Agentic & HumanintheLoop Design
- Define and orchestrate AI agents such as:
- Ingestion, extraction, context, and resultant calculation agents
- Stackup analysis agents for 2D and 3D assemblies
- Implement humanintheloop workflows allowing engineers to review, correct, and approve intermediate results
- Design traceability, audit trails, and explainability into AI decisions
AI/ML & GenAI Architecture
- Select and design deployment of SLMs/LLMs suitable for onprem servers (e.g., Llama, Mistral, Qwen, Phi3)
- Architect:
- Embeddings and vector search (FAISS / OpenSearch)
- Prompt and systemprompt strategies
- Optional finetuning workflows where applicable
- Ensure deterministic, engineeringsafe behavior for critical calculations
Technology & Integration Leadership
- Define integration patterns with:
- CAD/PLM systems (CATIA, Autodesk, PLM tools)
- ERP/QIS systems (for future QN integrations)
- Guide backend, API, and frontend teams using:
- FastAPI / REST
- ReactJS
- Kubernetes / Airflow orchestration
- PostgreSQL / MongoDB storage
NonFunctional Ownership
- Own performance, scalability, and reliability targets for large assemblies and datasets
- Work with IT teams on GPU/CPU sizing and infrastructure planning for onprem AI workloads
- Ensure production readiness, including security, OAuth2based access control, and deployment packaging
Required Skills & Experience
Core Expertise
- 12+ years in AI/ML or advanced engineering software architecture
- Strong handson experience with mechanical tolerance analysis / stackup analysis in aerospace or complex manufacturing domains
- Proven experience delivering onpremise AI platforms in regulated or IPsensitive environments
AI & Data
- Deep understanding of:
- GenAI frameworks (LangChain, LangGraph, Semantic Kernel, CrewAI)
- Vector databases and embeddings
- AI agent orchestration patterns
- Ability to combine deterministic mathematical models with probabilistic AI outputs
Engineering Systems
- Strong exposure to:
- CATIA, Autodesk CAD, PLM systems
- Reading and interpreting engineering drawings and GD&T
- Experience integrating AI systems with engineering workflows and SMEs
Software & Platform
- Strong backend and systemdesign skills (Python, C++, NodeJS)
- Experience with microservices, APIs, Kubernetes, and enterprisegrade web applications
Nice to Have
- Aerospace or engine system experience
- Experience with quality systems or downstream manufacturing decision support
- Prior work on AI explainability or validation for safetycritical systems
