Roles and Responsibilities

• Architect the Hybrid AI Vision: Lead the architectural design and strategic vision for hybrid inference systems, dynamically distributing Large Language Model (LLM) and Vision-Language Model (VLM) workloads across edge computing environments and cloud infrastructure.
• Team Leadership & Innovation: Lead, mentor, and inspire a team of specialized engineers working across distributed systems orchestration, inference optimization, and AI compiler engineering. While you are not expected to be a hands-on master of every domain, you will drive the overarching technical roadmap, foster a culture of cutting-edge innovation, and guide domain experts in navigating complex system tradeoffs.
• Design Dynamic Orchestration & Resilience: Oversee the architecture of high-availability orchestration engines that intelligently route inference tasks. Guide the team in developing cascading inference mechanisms, dynamic model fallback strategies, and robust telemetry to ensure continuous, steady-state inference under varying connectivity constraints.

Qualifications

• Education & Experience: Ph.D. in Computer Science, Computer Engineering, Artificial Intelligence, or a related field with 8+ years of relevant industry experience (or Master’s degree with 12+ years), including proven experience leading technical teams or driving complex architectural roadmaps.
• End-to-End Systems Leadership (T-Shaped Profile): Demonstrated capability to lead full-stack AI systems engineering. You possess deep, hands-on mastery in at least one or two of the following core domains, coupled with the comprehensive systemic breadth required to effectively lead engineers working across the others:
  • Distributed Systems & Hybrid Inference: Designing, scaling, and deploying production-grade distributed ML systems. Balancing cloud infrastructure with edge constraints using modern routing paradigms, such as cascading inference architectures and semantic routing.
  • Algorithmic & Inference Optimization: Proven experience optimizing state-of-the-art LLM/VLM inference pipelines. Deep understanding of model compression (e.g., PTQ, QAT, AWQ, FP8/INT4), hardware-aware compute optimizations (e.g., FlashAttention), and advanced memory management (e.g., PagedAttention, KV cache compression/eviction).
  • Advanced Systems & Compiler Engineering: C++ and production-grade Python proficiency. Deep understanding of edge/cloud model-serving frameworks (e.g., vLLM, TensorRT-LLM, ExecuTorch, MLC-LLM) and AI compilers (e.g., MLIR, Apache TVM, Triton) for compute graph optimization and custom kernel development.

Preferred Qualifications

• Privacy & Security: Deep understanding of privacy-preserving AI techniques (federated learning, differential privacy, secure enclaves) essential for processing sensitive data across edge and cloud environments.
• Community Engagement & Open Source: Publications in relevant AI, ML, or systems conferences (e.g., NeurIPS, ICML, MLSys), or active contributions to open-source ML infrastructure projects (e.g., vLLM, ONNX Runtime, Apache TVM, llama.cpp).