Siddharth Yadav — AI Operations, LLMOps, Platform Engineering
scroll down ↓ · $ cat resume.pdf
about
Siddharth Yadav
AI Operations · LLMOps · Platform Engineering · Cloud
AI Operations and Platform Engineering leader with 13+ years building the infrastructure that makes AI work in production. Most recently architected enterprise LLM platforms at McKinsey's $25B AUM investment arm — deploying Amazon Bedrock, Claude 3.5 Sonnet, and ChatGPT Enterprise to automate 10,000+ pages of documentation synthesis and reduce MTTR by 60% via AI-driven ops.
Before AI, the day job was scale: I led the modernization of 150+ legacy applications to multi-cloud (AWS + Azure), built the AppSec program from zero, and delivered a disaster recovery system in one-third the time AWS ProServe estimated. Now finishing MIT Applied AI & Data Science — in progress, expected May 2026 — and looking for what's next at the intersection of LLM operations and platform engineering.
experience
- Owned platform and security infrastructure powering a $25B AUM in-house hedge fund
- Delivered AWS-native Disaster Recovery in 1/3 the time at half the cost of AWS ProServe's $3.5M estimate
- Architected Enterprise AI Platform using Amazon Bedrock and Claude 3.5 Sonnet — automated synthesis of 10,000+ pages of internal documentation
- Deployed AI assistant that reduced MTTR by 60% via correlation of architecture docs with live infrastructure schemas
- Built Secure GenAI Research Assistant for Quants using Amazon Bedrock + SageMaker JumpStart
- Led enterprise rollout of ChatGPT Enterprise, Amazon Q, and Windsurf — secure onboarding of 100+ developers to LLM-powered coding and productivity tools
- Led organization restructuring of Database and IT Operations teams — $2M annual cost savings
- Spearheaded AppSec program — implemented SAST (Checkmarx), SCA (Blackduck), and security rating framework
- Architected and deployed AWS WorkSpaces (Linux + Windows) for high-frequency traders and investment professionals — delivering low-latency remote desktop infrastructure for mission-critical trading workflows
- Built automated FinOps framework to decommission AWS resources during off-hours — reducing cloud spend through scheduled scaling and resource lifecycle management
- Managed perimeter security with Cisco and Palo Alto firewalls across hybrid environments
- Designed and implemented DevOps, Cloud, and Security frameworks from the ground up — building MIO's platform engineering practice
- Modernized and containerized 150+ legacy applications (Python, C#, Java, Node) from monolithic VMs to OpenShift and Docker
- Orchestrated multi-cloud migration of 150+ apps to AWS and Azure (ECS, Fargate, AKS, Lambda, Step Functions) — $500K annual recurring savings
- Implemented IaC across AWS and Azure using Terraform and CDK for consistent and reproducible deployments
- Planned shutdown and decommissioning of two physical data centers (Chicago and Mahwah) with zero downtime
- Built and trained high-performing engineering teams across Jenkins, GitLab, GitHub, Docker, and OpenShift
Consultant for HSBC & Mastercard via Open Systems
- Built Azure-based Elasticsearch stack for a search engine prototype
- Automated app build and deployment using Kubernetes and Docker
- Established Spinnaker-based multi-cloud continuous delivery platform
- Led redesign of CI/CD strategy using ARM Templates and Terraform
Short-term contract via Consultnet
- Implemented Jenkins, Docker, and Artifactory platform — reduced app onboarding time by 80%, saved $400K
- Optimized build infrastructure using scalable Docker slaves — $100K cost reduction
- Led and trained a team of DevOps engineers
- Integrated 30 internal tools, 11 vendor tools, and 42 data sources into a unified technology management platform
- Built centralized Cloudbees Jenkins infrastructure enabling CI/CD across the enterprise
- Developed pipeline-as-code shared libraries supporting build, test, and deployment across technologies
- Automated infrastructure provisioning using Salt, Puppet, and Ansible
- Led mobile build and deployment for iOS and Android apps integrated with Jenkins and TFS
- Reduced production support resource engagement by 50% — $150K annual savings
- Created Data Dictionary baseline for the Data Architecture group
- Analyzed logistics and inventory management across 8 distribution centers
- Designed systems in Troux for TOGAF enterprise architecture
- Completed Six Sigma and IT Project Management training
projects
AWS Disaster Recovery Architecture
Designed and delivered a mission-critical AWS-native Disaster Recovery system for a $25B AUM hedge fund. Completed in 1/3 the time and at half the cost compared to AWS ProServe's $3.5M estimate.
1/3 delivery time · $1.75M under estimate
Enterprise AI Platform (Amazon Bedrock + Claude)
Architected an enterprise AI platform strategy leveraging Amazon Bedrock and Claude 3.5 Sonnet. Automated synthesis of 10,000+ pages of internal technical documentation and deployed an AI assistant that correlated architecture docs with live infrastructure schemas.
60% MTTR reduction · 10K+ pages automated
Serverless Quant Alpha Research Pipeline
End-to-end serverless ML pipeline for quantitative research — SageMaker Notebooks for model development, SageMaker Processing jobs for large-scale data analysis of unstructured market data.
Enabled quant researchers to run production-grade ML experiments without infrastructure overhead
150+ Application Cloud Modernization
Led end-to-end modernization and containerization of 150+ legacy applications (Python, C#, Java, Node.js) from monolithic VM and physical hardware environments to a multi-cloud setup spanning AWS and Azure.
$500K annual savings · zero downtime migration
Secure GenAI Research Assistant for Quants
Developed a secure LLM-based research assistant for quantitative analysts to explore unstructured market data. Built on Amazon Bedrock and SageMaker JumpStart with a serverless alpha research pipeline.
Reduced manual research time by hours per analyst per day
Application Security Program
Spearheaded the development of a company-wide AppSec program from the ground up — including a security rating framework, SAST tooling (Checkmarx), SCA tooling (Blackduck), and AWS Service Control Policies.
First formal AppSec program at MIO Partners
Data Center Decommissioning
Planned and executed the full decommissioning of two physical data centers in Chicago and Mahwah. Migrated all applications and technology assets to multi-cloud with zero service disruption.
2 data centers eliminated · significant CapEx reduction
Jenkins Platform at Broadridge
Implemented an open-source Jenkins, Docker, and Artifactory platform. Reduced app onboarding time by 80% through scalable Docker build slaves and automated pipelines.
80% faster onboarding · $500K total savings
Enterprise CI/CD Transformation at BofA
Unified 30 internal tools, 11 vendor tools, and 42 data sources into a centralized technology management platform. Built pipeline-as-code shared libraries and automated database deployments across SQL Server, Oracle, and IBM DB2.
50% reduction in production support · $150K annual savings
Personal LLM Projects // coming soon
LLM-powered applications and experiments built outside of work — details coming soon.
Exploring applied AI, RAG architectures, and LLM ops patterns
skills
AI Operations & LLMOps
Cloud
DevOps & CI/CD
Security
Databases
Languages & Frameworks
Observability
Leadership
education
in progress · expected May 2026 Certificate in Applied AI and Data Science
MIT Professional Education
2011 – 2013 Master of Science, Information Systems
University of Florida
Publicity Coordinator, SPICMACAY-UF; Member, Association for Information Systems
2007 – 2011 Bachelor of Technology, Computer Science Engineering
VIT University
contact
$ cat contact.txt