JPMorgan Chase
Lead Software Engineer (AI Platform) - Vice President
$157K–$215K
New York, NY · Director · Full-time · Posted Oct 1, 2026
- Experience
- At least 4 years of relevant industry experience.
- Education
- Master's degree in computer science, software engineering, computer engineering, information systems, or a related technical field.
About the job
Engineers and operates secure, scalable cloud and compute platforms that enable AI researchers to develop, evaluate, and prepare AI solutions for downstream integration.
What you'll do
- Own the AI research platform’s architecture, roadmap, and day-to-day engineering, including cloud and CPU/GPU infrastructure.
- Build reusable, automated research environments and platform services; translate AI research needs into dependable capabilities and support prototype integration.
- Maintain platform reliability and controls, and assess new AI infrastructure technologies through benchmarks, reviews, and pilots.
What they ask for
- Master's degree in a related technical field and at least 4 years of relevant industry experience building or operating cloud, platform, or infrastructure systems.
- Hands-on production environment experience, including EC2, networking, storage, monitoring, and cost management; knowledge of Linux, containers, and distributed systems.
- Experience with infrastructure-as-code, configuration management, and CI/CD; proficiency in Python and shell scripting.
- Working knowledge of AI/ML workflows and ability to translate research requirements into platform designs.
Nice to have
- Experience supporting AI/ML teams or GPU-intensive, LLM, agentic, or distributed AI workloads.
- Experience with Kubernetes, GPU software stacks, Slurm or Ray, observability platforms, cloud cost optimization, or regulated enterprises.
Pay
USD 156,750.00-$215,000.00 base salary
Tools
EC2, Terraform, CloudFormation, Python, Linux, Kubernetes, Slurm, Ray
AI focus
Enables AI research workloads, including LLM inference and evaluation, agentic workflows, retrieval, model serving, and accelerator benchmarking.
Model-written summary · the original posting is authoritative