Ayush Shukla

About Ayush Shukla

I have 6.5 years of professional experience, with 5+ years focused on cloud infrastructure, DevOps, and Agentic AI across enterprise clients in pharma, insurance, retail, and healthcare.My recent work centers around building Agentic AI solutions: Case creation Agentic AI solution that expanded into live Health Insurance IVR; voice-to-voice Agentic IVR (AWS Lex + OpenAI) with self-learning pipelines; and a ticket-classification pipeline tickets, pattern-based root cause and resolution) contributing to $1.6B in admin and cost-of-care impact, alongside Deepagents for drug-code validation and appeal letters (~20% time/cost reduction).On AWS infrastructure automation and cost optimization: Terraform and Jenkins automations that cut manual effort by more than 95%, and cost savings of $59k+ per month through EC2 stop/start, snapshot cleanup, rightsizing, and decommissioning resources—plus Windows Server 2022 Core migration and reliability improvements. I started in pharmacovigilance and cloud migration: leading backend design and implementation for Oracle Argus Safety upgrades and AWS migrations (workflow automation, SQL, DITL/UAT, zero-defect go-lives), then supporting multi-app AWS estates—patching 300+ servers, Terraform/CloudFormation provisioning, application upgrades, EBS and RDS hardening. From there I moved into DevOps and cloud operations for retail: deployments across 110+ servers, PagerDuty on-call, Lambda/Terraform, IoT dashboards, and disciplined release tracking.I am passionate about delivering measurable outcomes—reduced cost, higher reliability, and AI that moves from POC to production. I work with AWS, Terraform, Jenkins, Python, Docker and the OpenAI/LangChain/LangGraph stack and am always looking to connect with teams building cloud-native and AI-powered solutions.

Experience

1. Consultant Cloud Engineering

   Deloitte

   Jun 2023 — Present · Bengaluru, IN

   Cloud Engineer driving DevOps, infra automation, cost optimization, and Agentic AI for retail and healthcare clients.Retail DevOps: 8+ deployments across 110+ servers (75 Live, 22 LO, 10 Smoke-test); 5+ store onboardings; RKE2 on 10+ servers. On-call—100+ PagerDuty alerts; 30+ handovers, health-check and low-perf reports, weekly app-availability analysis; Lambda/Terraform upgrades; IoT dashboard; 30+ JIRA BTPs weekly; 6+ bi-weekly post-deploy validations.AWS Infra & Cost Optimization:$59k+ monthly savings via automations and rightsizing across 3 AWS accounts. Automations: EC2 stop/start (Lambda, CloudBees, EventBridge) on 20+ servers—1,959+ hrs/week avoided; snapshot cleanup $4,265/mo; self-service scheduler—50% admin cut. Rightsizing:$54,434/mo (30+ EC2, 238+ EBS, 30+ Fsx, instance rightsizing); 30+ Windows→2022 Core AMI (15–20% reduction). Terraform + Jenkins (weeks→hours, 6+ envs);>95% manual reduction; Terraform Destroy fix, WinRM/UCD ~20% faster. Agentic AI (IVR & case creation): 8-week Agentic AI POC (OpenAI, LangChain, LangGraph, SageMaker, Python) for case creation—replacing manual patient history/eligibility review with AI-driven decisions; POC expanded to live Health Insurance IVR. Building voice-to-voice Agentic IVR (AWS Lex + OpenAI): real-time responses for providers on claims, eligibility, benefits, case history; 3 intents, 15 sub-intents; training on case history and docs (BPDs, faxes); self-learning pipeline on live call data to cut fallback rates.Agentic AI (ticket classification & deepagents): Ticket Classification pipeline (Python, LangChain, OpenAI): tickets→structured tagging, patterns→root cause, investigation, resolution steps—$1.6B admin/care cost impact. Deepagents for drug-code validation and appeal letters (~20% time/cost reduction).~10k IVR transcripts (11+/55+ intents); NovaSonic TTS evaluation.

Education

• Jaypee Institute Of Information Technology

  Bachelor of Technology - BTech, Electronics and Communication Engineering

  2015 — 2019