About
I have spent the last several years at the deep end of container and AI infrastructure on AWS — Amazon EKS and ECS, GenAI inference on both NVIDIA GPUs and AWS silicon (Trainium / Inferentia), and the agentic-AI platforms enterprises are now building on top of them.
My work sits where architecture meets outcomes: cutting GPU inference cost and latency, keeping mission-critical banking platforms resilient, defending scalability at world-record streaming concurrency, and turning hard-won field patterns into workshops, blogs, open-source assets and reference guidance that other engineers can reuse.
By the numbers
What I do
AI infrastructure & GPU inference
LLM inference optimisation on Amazon EKS — tiered KV-cache offloading (vLLM, LMCache, ElastiCache) delivering up to 5× faster time-to-first-token and ~60% lower GPU cost; GPU autoscaling and cross-region GPU-scarcity failover; Ray and Anyscale on EKS; and deep Neuron / Trainium / Inferentia inference work.
Agentic AI platforms
Production agentic platforms on Kubernetes — AI gateways, LLM observability and evaluation, MCP tool servers, guardrails, and per-tenant AI cost governance (“tokenomics”).
Kubernetes & cloud-native at scale
Amazon EKS and ECS architecture for banking, fintech, OTT streaming and SaaS platforms — multi-tenancy, networking, resilience and disaster recovery, security and compliance baselines, upgrades at fleet scale, and serious cost optimisation.
Teaching & technical content
AWS workshops, Immersion Days, published blogs and case studies, open-source reference solutions, and conference talks — plus the courses and videos I create for engineers growing their cloud careers.
Connect
- LinkedIn — /in/jpalawala Professional updates, AI-infrastructure and Kubernetes posts.
- YouTube — @ctcwithjp Cloud, Kubernetes and AI lessons, and career guidance for engineers.
- Instagram — @cloudtechnologyconsulting Short-form lessons and behind-the-scenes.
- Email — alawalajp@gmail.com Speaking, collaboration and professional enquiries.
My CV
- CV — concise The two-minute read: summary, expertise, certifications, experience.
- CV — expanded, with links to every blog, workshop and repository For anyone who wants the detail and the verifiable artifacts.
- Download the PDF Three pages, print-ready.