Head of Spatial AI

About the Role

BitsBody develops next-generation computational modeling & simulation technologies to optimize engineering design & analysis workflows for spatial AI applications. We're seeking an entrepreneurial, high-ownership individual contributor to drive the core engineering, development, and deployment of our Generative AI platform, HumBoAI, an integral component of our medical technology solutions for anatomical modeling & simulation (M&S). This role is ideal for a hands-on direct execution operator with a proven ability to design, train, and productionize multi-modal generative machine learning (ML) pipelines, build robust backend infrastructure & APIs, and develop an interactive 3D web interface that visualizes model outputs in real-time.

Responsibilities

• Architect and optimize multi-modal, cross-domain end-to-end pipelines by training multi-scale vision and spatial-temporal foundational models, integrating language-model conditioning where appropriate.
• Containerize models and expose robust, low-latency APIs for heavy generative inference workloads.
• Productionize models as cloud services; implement model versioning, CI/CD, automated testing, and cloud-native deployment pipelines.
• Implement post-deployment monitoring, drift detection, A/B testing, and automated retraining workflows to maintain performance and safety.
• Implement and validate geometry pipelines (meshes, SDFs, implicit fields), ensure simulation-readiness, and integrate downstream finite element analysis (FEA) workflows.
• Architect and build a responsive, high-performance 3D web frontend and interactive workspace that visualizes real-time ML computations.
• Embed volumetric viewers and mesh editors with analysis toolsets for both medical images and polygonal meshes.
• Collaborate closely with the technical team to verify system-level behavior and translate core anatomical M&S research into stable, production-ready enterprise applications.
• Build early engineering frameworks and establish the architectural blueprint to recruit and lead the core spatial AI engineering team once institutional funding scales.

Required Qualifications

• Bachelor's degree (B.S./B.E./B.Tech) in Computer Science, Data Science, Machine Learning, Computer Vision, or a related quantitative or scientific field with at least 10 years of progressive experience; OR a master's degree (M.S./M.E./M.Tech) with at least 8 years of relevant experience; OR a doctoral degree (Ph.D.) with at least 4 years of relevant experience, demonstrating hands-on development of data-intensive web apps involving spatial AI & production-grade deployment capability in the required areas listed below.
• Experience building & training probabilistic generative models in 3D computer vision and spatial geometry (normalizing flows, diffusion models, VAEs, NeRF/implicit fields, and Gaussian Splatting) alongside natural language processing architectures (transformers, BERT, and T5) for handling image-to-3D or text-to-3D neural, generative pipelines.
• Expert-level command of high-performance computing with the following:
  • Backend/AI (Python for 3D ML): Proficiency in training graph neural networks and geometric deep learning systems using PyTorch ecosystem (including PyTorch3D, PyTorch Geometric, and Deep Graph Library), MONAI for medical imaging workflows, and familiarity with TensorFlow ecosystem where applicable (including TensorFlow3D, TensorFlow Graphics, TensorFlow GNN, and Graph Net).
  • Frontend/Graphics (JavaScript/TypeScript): Proficiency in real-time streaming of large volumes arrays & dense geometric meshes, including Three.js/React Three Fiber/Babylon.js, WebAssembly (Wasm), and WebGL/WebGPU.
  • Backend & APIs: Proven experience building robust APIs with FastAPI, GraphQL, gRPC/Protobuf (for efficient large 3D array transfers), and streaming protocols (WebSockets/WebRTC) to push progressive updates to the client; familiarity with Celery, Redis, PostgreSQL, and object storage for large 3D arrays.
  • Production DevOps: Deep experience containerizing heavy GPU workloads using Docker and orchestrating scalable deployments with Kubernetes and Helm on AWS or GCP (EKS/GKE, S3/Cloud Storage, IAM).
  • Model Serving & Inference: Experience with NVIDIA Triton, ONNX Runtime, TorchServe, or custom FastAPI/gRPC inference stacks; knowledge of model quantization, graph pruning, and inference optimization.
  • Performance Engineering: GPU programming familiarity (CUDA, cuDNN), distributed training (NCCL), and profiling/optimization for large models and volumetric data.
  • 3D Tooling & Data Libraries: Proficient with medical imaging frameworks (PyDicom, SimpleITK/ITK, NiBabel for DICOM/NIfTI preprocessing and segmentation) and 3D mesh/point cloud manipulation libraries (Open3D, Trimesh, MeshLib, Kaolin or equivalent).
• Practical understanding of PHI handling, DICOM security considerations, and HIPAA-aware engineering practices.
• Ability to operate autonomously in ambiguity and build structure from absolute zero.

Preferred Qualifications

• Prior experience within the early-stage startup lifecycle (founder or early employee as a technical leader in Pre-seed through Series A), demonstrating resilience & adaptability in a steep learning-focused environment.
• Experience compiling and optimizing C++ kernels for performance-critical pipelines – medical image volumes (e.g., ITK, VTK) and geometry (e.g., OpenCASCADE, Manifold3D).
• Experience with medical imaging pipelines (CT/MRI preprocessing, segmentation, registration), physics-based simulation engines (OpenSim, FEA integrations), and geometry/CAE toolchains (OpenCASCADE, VTK).
• Strong peer-reviewed publications or open-source contributions in ML, computer vision, or 3D geometry.

Compensation Pathways

All equity allocations or stock options are subject to a standard 4-year vesting schedule with a 1-year cliff, matching institutional venture practices. Two distinct pathways are offered for this role:

• Option #1 (Pre-funding): Join as Co-founder & Head of Spatial AI before raising our first round of outside capital. This is initially an equity-only role, offering a meaningful co-founder-level equity stake. Upon securing outside funding, there is a clear path to a balanced salary & benefits package.
• Option #2 (Post-funding): Join as Head of Spatial AI after we raise outside capital. Compensation will include a competitive salary & benefits package. Stock options will be commensurate with a senior-level hire. Candidates selecting Option #2 will be considered only if no qualified candidate is found for Option #1.

Culture & Environment

• Mission-driven, high-velocity, and high-ownership environment.
• Direct influence on product, architecture, and long-term direction.
• Opportunity to build a category-defining technology from the ground up.
• Culture of transparency, intellectual rigor, and responsible innovation.
• Collaborative environment bridging medical, engineering, science, and commercialization.
• Post-funding benefits will include health, dental, vision, retirement, and paid family leave.

This response is AI-generated, for reference only.