Sr Director of Software Engineering- AI Infrastructure Platform

JPMorganChase•Palo Alto, CA

About The Position

Your opportunity to make a real impact and shape the future of financial services is waiting for you. Let’s push the boundaries of what's possible together. As a Senior Director of Software Engineering at JPMorganChase within the firmwide AI Infrastructure Platform organization, you will lead multiple technical areas and manage the activities of multiple departments responsible for delivering a unified AI infrastructure layer across on‑premises environments, public cloud, and emerging accelerated‑compute vendors. You will collaborate across AI/ML engineering, infrastructure, security and controls, and vendor teams to ensure the firm remains at the forefront of AI platform capabilities, operational excellence, and industry best practices. In this role, you will own training and experimentation on a Kubernetes‑standardized platform. While a dedicated architecture function exists, you will act as an active design partner—guiding architectural trade‑offs and ensuring designs translate into reliable, secure, and operable systems at enterprise scale.

Requirements

15+ years of engineering experience, including 8+ years of senior engineering leadership experience with responsibility for managing managers.
Demonstrated experience delivering platform products (beyond foundational infrastructure) with strong adoption, reliability, and operational maturity.
Experience developing and leading large, cross‑functional engineering teams within highly matrixed and complex enterprise environments.
Proven track record of leading complex initiatives supporting distributed system design, testing, and operational stability at scale.
Deep hands‑on expertise with Kubernetes‑based platforms , including: Multi‑tenancy, RBAC, admission control, and network policy, Multi‑cluster operations, upgrades, and cluster lifecycle management, Controllers, operators (CRDs), and platform API design patterns
Experience supporting AI training and experimentation platforms, including: PyTorch and distributed training concepts such as scaling, orchestration, and failure modes, Ray or similar frameworks for distributed experimentation execution, Familiarity with Slurm or equivalent HPC or batch schedulers and core concepts such as queues, fair‑share, reservations, and preemption
Understanding of modern AI inference stacks (for example, vLLM) and how serving constraints—latency, throughput, batching, KV cache behavior, and GPU memory limits—influence training and experimentation platform design.
Strong understanding of GPU infrastructure fundamentals, including NVIDIA ecosystem capabilities, health and telemetry signals, and scheduling and placement constraints.
Extensive practical experience with cloud‑native technologies and hybrid infrastructure environments spanning on‑premises and public cloud.
Experience hiring, developing, coaching, and retaining high‑performing engineering talent.

Nice To Haves

Experience operating large‑scale GPU fleets, including heterogeneous accelerator environments.
Experience delivering hybrid AI platforms across on‑premises infrastructure, public cloud, and specialized accelerated‑compute vendors.
Experience working at the code level within large‑scale distributed systems.

Responsibilities

Lead multiple technology and platform implementations across departments to deliver firmwide AI infrastructure objectives, with a primary focus on training and experimentation platforms operating at enterprise scale.
Own the design, delivery, and evolution of a Kubernetes‑first training and experimentation platform , including Kubernetes‑native support for batch and distributed training jobs, lifecycle management, retry semantics, and failure recovery patterns.
Standardize AI developer workflows for experimentation, enabling self‑service job submission, reusable templates and golden paths, reproducibility mechanisms, and consistent runtime behavior across hybrid deployment environments.
Build and evolve platform APIs and automation, including Kubernetes controllers and operators where appropriate, to ensure the platform is safe, scalable, and easy to adopt across teams.
Drive measurable improvements in GPU availability and utilization through reliability engineering, fleet readiness patterns, and accelerated capacity onboarding.
Define and implement governance‑based scheduling and placement strategies, including: Multi‑tenant GPU quotas and guardrails, Priority, admission control, and reservation patterns, Preemption policies, Fragmentation reduction and topology‑aware placement (GPU type, MIG, and topology awareness)
Embed enterprise‑grade security, risk, and control requirements into platform defaults, including IAM and RBAC controls, secrets management, audit logging, policy enforcement, network segmentation, and controlled change management.
Drive operational excellence by establishing SLIs and SLOs, managing error budgets, leading incident management practices, forecasting capacity, and delivering end‑to‑end platform observability across job lifecycles and GPU telemetry.
Act as the primary interface with senior leaders, stakeholders, and executives, driving alignment and consensus across competing priorities and complex initiatives.
Lead multiple engineering teams and managers, building a high‑performing organization with strong engineering standards, scalable operating models, and a culture of accountability and continuous improvement.
Champion the firm’s culture of diversity, opportunity, inclusion, and respect.