Principal Architect · New York, NY

Cloud, GenAI,
and Database, built
for production.

Complex data, security, and architecture challenges are the ones I most want to help with. I work with enterprises to turn ambitious ideas into systems that actually run, across AWS, Google Cloud, and Microsoft Azure. Currently leading partner architecture at Neo4j, where graph data meets generative AI.

  • AWS
  • Google Cloud
  • Microsoft Azure
  • Databricks
  • Snowflake
  • Salesforce
  • Neo4j
Portrait of Guhan Sivaji
Principal Partner Architect, Neo4j
01

About

I am a Principal Solutions Architect with two decades across cloud data management, database platforms, and generative AI. My work sits at the intersection of three things that have to coexist cleanly: how data is modeled and governed, what database is right for the problem, and how AI can be applied without ignoring either.

At Neo4j I lead end-to-end solution architecture for enterprise AI deployments, bringing retrieval-augmented generation, vector search, and knowledge graphs into real customer environments. Before Neo4j, I spent more than eleven years at Informatica working on master data management, data integration, and data quality at scale, with earlier stops at Dell EMC, BD, Bank of America, Sun Microsystems, and GE.

Throughout that arc, the constant has been the data layer. Cloud is a substrate. Databases and the way data flows through them are where most architectures actually succeed or fail.

I care deeply about secure, well-architected systems. Zero-trust thinking, least privilege, identity-first design, and operational simplicity are non-negotiable for me, regardless of how exciting the model on top is.

I also spend a meaningful amount of my time on stage and in rooms full of practitioners. Frequent speaker at industry events and technical conferences, with hundreds of partner workshops, executive briefings, and architectural deep dives delivered across the AWS, Google Cloud, and Microsoft ecosystems.

02

How I work

Collaboration over heroics

I would rather move a team forward than be the smartest voice in the room. The smartest voice is almost never one voice.

Comfortable changing course

Plans should answer to evidence, not pride. I will redesign cheerfully when the data asks for it.

Same curiosity at every stage

Whether the conversation is about why something broke or what we are building next, I show up the same way.

Plain language, calm tone

Complex systems do not need complex talkers. I aim to leave the room less confused than I found it.

03

Experience

Nov 2022 – Present

Principal Partner Architect · Neo4j

Lead enterprise GenAI architecture for partners and customers across the AWS, Google Cloud, and Microsoft Azure ecosystems. Design production AI systems combining LLMs, vector search, and knowledge graphs for high-throughput inference and real-time decisioning. Serve as escalation architect for complex distributed-systems and AI-infrastructure engagements.

Aug 2011 – Dec 2022

Senior Cloud Data Management Specialist · Informatica

Eleven-plus years across pre-sales, professional services, and architecture leadership. Built and led C-suite engagements, authored technical proposals, and ran POC environments across AWS, Azure, and GCP. Managed delivery teams across financial services, healthcare, and manufacturing.

Nov 2010 – Aug 2011

Principal Consultant, MDM · Dell EMC

MDM implementations for Fortune 100 clients. Defined data governance processes, modeling standards, and matching rules alongside data stewards and analysts.

Earlier roles at BD, Bank of America / Merrill Lynch (Siperian), Sun Microsystems, and GE.

04

What I work on

Cloud Data Management

Master data management, data integration, data quality, and governance, hardened through more than a decade at Informatica. The unglamorous work that decides whether an AI program ever sees production.

  • MDM
  • Data Quality
  • Data Governance
  • iPaaS
  • ETL / ELT

Generative AI & LLMs

RAG, vector search, knowledge-graph grounding, agentic orchestration, prompt engineering, embedding pipelines, MCP integrations across Claude, Gemini, and agentic IDEs.

  • RAG
  • LangChain
  • LlamaIndex
  • Claude API
  • MCP

Cloud & Distributed Systems

Cloud-native, security-first systems on AWS, Azure, and GCP. High-throughput inference, container orchestration, IaC with Terraform and CloudFormation, DevSecOps in the pipeline.

  • AWS
  • GCP
  • Azure
  • Kubernetes
  • Terraform
  • Docker
05

Credentials

AWS Solutions Architect

Professional · Earned 2018

Google Cloud Architect

Professional · Earned 2024

Azure Solutions Architect

Expert · Earned 2020

Snowflake SnowPro

Core

Also certified: Neo4j Certified Professional, Neo4j Graph Data Science, Informatica MDM Multidomain, Career Essentials in Generative AI by Microsoft and LinkedIn, and Claude Code in Action by Anthropic (Apr 2026).

06

Awards

Google Cloud Partner All-Star, Sales

Google Cloud · 2023

Product Specialist of the Year

Informatica · 2022

President’s Club

Informatica · 2019, 2020

Also recognized: Winner, Databricks DNB Hackathon (2025); Presales Product Specialist of the Year, North America (2020); and Engineering Excellence Award, Sun Microsystems (2006).

07

Open Source

A selection of partner integration projects I have led or contributed to, bringing Neo4j into the cloud and SaaS ecosystems our customers actually run on.

Snowflake

cortex-code-neo4j-mcp

MCP server exposing Neo4j to Snowflake Cortex agents, so Cortex-side reasoning can query graph context directly.

Databricks

neo4j-databricks-agent

Databricks agent that brings Neo4j graph reasoning into the Databricks AI runtime and notebooks.

AWS

neo4j-aws-privatelink-443

AWS PrivateLink reference for Neo4j on port 443, the production pattern for keeping graph traffic off the public internet.

Azure

azure-databricks-aura-privatelink

Private connectivity blueprint between Azure Databricks and Neo4j Aura via Private Link, with end-to-end network isolation.

ServiceNow

servicenow-neo4j-kafka-cdc

Streams ServiceNow change events into Neo4j through Kafka CDC for real-time CMDB and incident graphs.

Google Cloud

neo4j-aura-gcp-psc

Neo4j Aura on Google Cloud over Private Service Connect, deployable as a clean, secure pattern for enterprise tenants.

Slack

neo4j-slack-bot

Slack bot that talks to a Neo4j knowledge graph, so teams can query their own data without leaving the channel.

Fraud Detection

neo4j-fraud-detection

Reference architecture for graph-powered fraud detection, with the queries and patterns that actually catch the ring structures.

08

Writing

A selection of articles I have authored or contributed to on graph databases, GenAI, and cloud data platforms.

09

Off the clock

Woodworking

Tables, a loft bed, smaller projects in between. Hand tools, joinery, the smell of fresh shavings. A counterweight to a life spent in front of screens.

Automobiles

Hands-on, not just curious. Spark plugs, fluids, brake rotors, the occasional air intake manifold that takes longer than it should. The kind of weekend project that teaches you something every time.

Photography

Mostly travel and family, occasionally the dog when he sits still long enough.

Restoration videos

Old cars, old machines, old tools. The patience of bringing something forgotten back to life never gets old.

Pets

One dog and one cat. They run the household. I just contribute the salary.

10

Get in touch

Whether you are exploring how graph and generative AI can transform your business, or you need a second set of eyes on an architecture decision, I am happy to talk. Grab a quick call, or say hi on LinkedIn.