ex-Google · ex-Amazon · Austin, TX

Guru
Krish

Director — Enterprise AI Transformation, Governance & Security

$700M+

Business impact enabled

150+

Business units governed

18 yrs

Google & Amazon

7→1wk

AI pipeline delivery

gurukrish81@gmail.com LinkedIn

"I build the operating infrastructure that makes enterprise AI scalable, trustworthy, and governable — not by building the models, but by building everything around them."

What I do

Enterprise AI Transformation

Define how organizations adopt AI — strategy, platform infrastructure, operating model, and executive reporting.

AI Governance & Security

Build the controls, frameworks, and enforcement mechanisms that make AI safe to deploy at scale.

Hands-on AI Systems

Build and deploy LLM pipelines, governance workflows, and agentic systems — not just strategy documents.

NIST AI RMF AWS Bedrock Trust by Design LLM Governance Agentic AI

How I Think

Three principles that guide every engagement

"Governance that nobody uses is just overhead. The gateway has to be easier than the workaround."

At Google, Risk Inspector succeeded because compliance became the path of least resistance — not because we mandated it. At Amazon, adoption followed when we removed friction, not when we added policy.

"The platform scales. The headcount doesn't have to."

A centralized LLM gateway governing 600 engineers costs the same to operate whether you have 3 solutions engineers or 18. Build the mechanism first, then scale the team into it.

"Fix the pain before you ask for adoption."

7-week access backlogs at Amazon. 3-month privacy reviews at Google. In both cases, adoption followed from removing friction — not from training sessions or mandates.

Case Studies

Real programs. Real outcomes.

Google — 3P Security & Privacy

Risk Inspector: Governing 150+ Business Units at Scale

Walked into a company-wide Salesforce crisis — $700M in platform value blocked overnight. Designed 12 security and privacy controls, built Risk Inspector as the mandatory governance platform, and unblocked the ecosystem.

150d→10h Partner onboarding time

$500M+ Ecosystem value unlocked

3,700+ Licenses eliminated

Amazon — Security Data Intelligence

Security Scorecard: From Insight to V2 Launch

Owned the Security Scorecard product strategy — interviewed customers, demoed the product, analyzed usage via log queries, and synthesized findings into the v2 strategy adopted by engineering. Deliverables fed into OP1 and OP2 planning cycles.

Phase 1 Product strategy delivered and shipped

25% Usage increase post-launch

OP1+2 Planning cycles informed

Google — AI/ML Strategy

Enterprise AI Operating Model: 14 Verticals, $50M

Built and ran the AI/ML strategy program across 14 Google corporate verticals — intake framework, financial impact model with Finance, SteerCo with 20+ Directors and SVPs, OKR alignment across 50+ programs.

50+ AI programs governed

$30-50M Portfolio managed

14 Corporate verticals

Frameworks & Playbooks

Reusable architectures for enterprise AI

Governance Framework

Trust by Design

A reusable control architecture for enterprise AI systems — risk tiering, agentic AI threat models, guardrail enforcement patterns, and evaluation frameworks for safe LLM deployment. Built from 18 years of governance experience at Google and Amazon.

View on GitHub

Transformation Playbook

Enterprise AI Transformation Playbook

From early traction to durable scale — the operating model, platform architecture, intake framework, and scaling flywheel that makes solution #50 as governed and reliable as solution #5. Coming soon as a published document.

Request a copy

Working AI Systems

Built, not just designed

Research & Applied — Amazon

LLM Data Access Pipeline — Amazon

Built an LLM classification pipeline on AWS Bedrock to analyze security access request tickets — Python API to extract ticket data, stored in S3, classified in Jupyter Notebook using prompt engineering to identify what data customers were requesting and why. Used findings to: build new features in the Shepherd security product, create a new data access policy published internally, design an optimized data access request workflow, and reduce engineering oncall burden. Ran weekly office hours, trained the data team oncall, personally took ownership of access approvals during transition. Resolution time: 7 weeks → 1 week.

AWS Bedrock Python S3 Jupyter Prompt Engineering Shepherd Security

In Progress

AI Governance Intake Classifier

Takes a business AI use case description, classifies risk tier (Low/Medium/High/Critical), generates discovery questions, recommends whether to proceed and who needs to approve. Built in n8n with LLM reasoning layer.

n8n Claude API LangChain

→ Live demo coming soon

In Progress

Agentic AI Research Workflow

Multi-step research agent that takes a governance question, retrieves relevant information from a knowledge base, synthesizes findings, and returns a structured recommendation with sources and confidence level.

LangGraph Claude API RAG Vector DB