Senior AI Architect - GenAI & Enterprise AI Leader
Resume:
ARUNA KANAPARTHY
Senior AI Architect GenAI & ML Solutions Leader
20+ Years · GCP · Vertex AI · RAG · LLMs · Agentic AI
*****.*******@*****.***
linkedin.com/in/aruna-kanaparthy-a9917a4/
Dallas-Fort Worth, TX
AI & TECHNICAL EXPERTISE
Generative AI
RAG · LLMs ·
Fine-tuning
Agentic AI
ADK · LangChain ·
ReAct
Vertex AI
Training · Inference ·
MLOps
Cloud Platforms
GCP · AWS · Azure
Enterprise Search
RAG · CCAI · Retail
Search
Conversational AI
Dialogflow · Agent Assist
Data Engineering
Dataflow · Composer ·
ETL
Solution Architecture
Presales · POCs · SOWs
ML Pipelines
CI/CD/CT · Kubeflow
Industry AI
Healthcare · FinServ ·
Retail
EXECUTIVE SUMMARY
Hands-on AI Architect and technology leader with 20+ years designing and deploying production-grade AI systems across enterprise environments. Deep practitioner expertise in Generative AI — including RAG architecture, LLM fine-tuning, agentic frameworks, and multimodal solutions — combined with the strategic vision to translate complex business challenges into scalable AI platforms. Proven track record of owning AI solutions end-to-end: from requirement discovery and solution architecture, through hands-on implementation on GCP (Vertex AI), AWS, and Azure, to production deployment and team enablement. Recognized thought leader who builds reusable AI accelerators, runs internal hackathons, and coaches teams on applied GenAI best practices. Currently driving enterprise AI adoption at CGI while leading the GenAI Kitchen platform — an industry-specific GenAI solutions hub. PROFESSIONAL EXPERIENCE
Director, AI Thought Leadership & Architecture CGI Nov 2025 – Present Leading enterprise AI initiative design and development; building reusable AI accelerators and industry-specific AI demo platforms to accelerate go-to-market.
Architected and developed AI Launchpad, a voice-enabled autonomous AI interview and discovery platform leveraging Google Gemini Live APIs, capable of conducting multi-turn stakeholder interviews, dynamically adapting questioning strategies, and automatically extracting, synthesizing, and prioritizing AI use cases in real time using NLP pipelines, LLMs, and agentic orchestration frameworks — reducing enterprise discovery and requirements gathering time by more than 70%.
Designed and implemented production-grade agentic AI systems featuring conversational memory, planning, tool-calling, retrieval orchestration, multimodal reasoning, and adaptive workflow execution to support enterprise-scale AI discovery and decision intelligence workflows.
Engineered industry-specific AI simulation and demo environments across Oil & Gas, Healthcare, Financial Services, Retail, and Manufacturing domains using synthetic and anonymized enterprise datasets; built end-to-end GenAI demonstration flows to validate AI feasibility, operational value, and transformation opportunities during executive and presales engagements.
Built reusable enterprise AI accelerators integrating Vertex AI Agents, RAG architectures, semantic retrieval pipelines, streaming LLM responses, and contextual grounding frameworks to rapidly prototype and demonstrate scalable AI solutions while shortening enterprise sales and solution validation cycles.
Led hands-on architecture and implementation of cloud-native AI platforms using GCP Vertex AI, BigQuery, Cloud Run, FastAPI, WebSockets, Python, and React, enabling scalable real-time AI interactions, analytics, and orchestration capabilities.
Defined AI product strategy, technical roadmaps, architecture standards, and agile delivery backlogs in collaboration with business stakeholders, engineering teams, and leadership to accelerate iterative delivery of enterprise AI platform capabilities.
Delivered enterprise AI thought leadership through architecture patterns, internal enablement sessions, reusable reference implementations, and governance-focused discussions centered on Generative AI, agentic systems, responsible AI, evaluation frameworks, grounding strategies, and enterprise AI adoption best practices. Founder & AI Platform Architect GenAI Kitchen (Independent Platform) Sep 2024 – Present Building a multi-industry GenAI solutions platform bridging AI technology and real-world business applications, with comprehensive developer guides and working implementations.
Designed and implemented GenAI solution blueprints for Healthcare (clinical summarization, patient triage), Finance (risk analysis, regulatory Q&A), Retail (product search, recommendation engines), and Gaming (procedural content, personalization) verticals
Built scalable enterprise RAG architectures supporting large-scale semantic retrieval across structured and unstructured enterprise knowledge sources (Vertex AI Search, Embeddings API, Vector Search), including chunking strategies, embedding pipelines, retrieval ranking, and grounded generation
Developed hands-on code libraries and technical guides covering prompt engineering, tool-calling agents, multi-turn dialogue systems, and fine-tuned model deployment across major cloud platforms
Created AI educational content, interactive demos, and case studies translating complex GenAI concepts for both C-suite business leaders and engineering teams
Established partner relationships with enterprise clients to deliver customized AI solutions; led discovery workshops and architecture reviews to define AI adoption roadmaps AI/ML Architect — GCP Solutions CloudWerx Mar 2024 – Sep 2024 Hands-on AI architect designing and deploying transformative ML and GenAI solutions on Google Cloud Platform across six industry verticals.
Designed production-grade RAG architectures optimized for latency, grounding accuracy, retrieval quality, and enterprise scalability using Vertex AI Agents, LangChain, and ReAct frameworks; implemented semantic chunking, hybrid retrieval (dense + sparse), re-ranking, and grounded generation pipelines achieving high accuracy in enterprise knowledge retrieval
Architected fine-tuned BERT models and LLM solutions for text summarization, named entity extraction, sentiment classification, and context-aware predictions — improved NLP accuracy by 30%+ over baseline in healthcare and financial use cases
Built Conversational AI (CCAI) platforms using Vertex AI Agents and Dialogflow CX with context-aware, multi-turn intent handling, dynamic webhook fulfillment, and seamless live-agent escalation
Life Sciences & Healthcare: Designed AI-driven patient history summarization pipeline (structured + unstructured EHR data), predictive models for visit default risk, and semantic search over clinical documents
Life Sciences: Proposed and scoped DNA/RNA sequencing analysis automation leveraging Generative AI, Nextflow, and HPC (high-performance computing) integration on GCP
Gaming & Entertainment: Architected multimodal content generation pipelines (text-to-image, procedural narrative) and ML-based personalization models that boosted user engagement KPIs
Retail: Implemented AI product recommendation systems, customer behavior analytics (propensity modeling, churn prediction), and automated support using Vertex AI and GCP data stack
Real Estate: Delivered real-time data augmentation pipeline combined with traditional ML (XGBoost) and MLOps
(Vertex AI Pipelines) for automated property price estimation
Led internal hackathons driving rapid proof-of-concept development; partnered with Google to establish GCP GenAI implementation best practices across Vertex AI, CCAI, and RAG tooling
Spearheaded presales activities: requirements gathering, scoping sessions, SOW authoring, and solution architecture presentations for mid-to-enterprise clients Cloud AI/ML & Enterprise Search Architect SADA (Google Cloud Premier Partner) Dec 2018 – Feb 2024
Led design and delivery of high-impact Google Cloud AI solutions across search, analytics, ML pipelines, and early Generative AI POCs for enterprise clients.
Designed and implemented Retrieval-Augmented Generation (RAG) frameworks powering a BI Concierge solution
— enabling natural-language querying over enterprise data warehouses (BigQuery) using LLMs and grounding pipelines
Built automated AI-driven ad campaign generation system using GenAI + social media APIs; reduced campaign creation time by 60% while increasing creative output relevance
Architected Google Cloud Search enterprise deployments with custom connectors for Jira, Confluence, and ServiceNow — enabling unified intelligent search across enterprise knowledge silos for clients
Architected production ML pipelines on Vertex AI: model training, hyperparameter tuning, model registry, endpoint deployment, and CI/CD/CT automation with continuous evaluation
Operationalized enterprise AI pipelines on Vertex AI with CI/CD/CT, model monitoring, evaluation frameworks, and governance controls
Developed POCs for Retail Search and Recommendations AI prior to official product GA; played key role in validating these GCP products for production readiness
Incubated Looker practice within SADA — achieved delivery-verified status, built LookML solutions, and established data modeling best practices for the team
Architected CCAI solutions including real-time COVID-19 information bot (Dialogflow + Node.js), Agent Assist implementations, and Insights analytics for contact centers
Engaged in presales cycles spanning requirement gathering, solution architecture, SOW development, and technical presentations; contributed to millions in ACV for GCP-based AI engagements Solutions Architect — Enterprise Search & AEM Perficient Inc. Aug 2015 – Oct 2018
Led FedEx search relevancy optimization using Google Search Appliance (GSA) — architected entity recognition, custom collections, and metadata proxy servlets that improved search conversion rates
Developed custom AEM + GSA integration architecture for enterprise clients; led Children's Healthcare of Atlanta custom UI development as technical lead from requirements to delivery
Managed AEM 5.6 AEM 6 migration including code refactoring, performance optimization, and deprecated API remediation for large-scale enterprise web platforms EARLIER CAREER (2000–2015)
Broad enterprise technology foundation spanning ECM, content workflows, and Java/J2EE development:
D.R. Horton (2014–2015): Led Documentum/Captiva enterprise content migration; automated imaging workflows using Java and J2EE
SPE (2006–2013): Managed multi-organization digital library of 200K+ technical papers; integrated Google Search with Documentum ECM systems
JP Morgan Chase (2006): Built complex investor reporting workflows using Documentum ECM and J2EE/Struts; enterprise content routing and approval automation
Blue Shield of California (2003): Implemented Document Content Assembly (DCA) automation on Documentum + Oracle; designed PL/SQL procedures for XML-based contract generation
Infosys Technologies (2000–2001): Insurance systems development in PL/SQL and C; full lifecycle testing and module integration
SIGNATURE AI PROJECTS & IMPLEMENTATIONS
AI Launchpad
Voice-enabled AI interview system for automated use-case extraction. Built with streaming STT, LLM-powered entity extraction, and dynamic use-case synthesis pipeline. Reduces discovery workshops from days to hours.
BI Concierge (RAG)
Natural-language analytics interface over BigQuery using RAG + LLM grounding. Designed semantic retrieval layer, SQL generation pipeline, and hallucination mitigation via retrieval verification and source attribution.
Clinical AI
(Healthcare)
Patient history summarization system on GCP combining structured EHR data with unstructured clinical notes. Integrated Vertex AI for multi-document summarization and predictive visit default modeling using ML inference pipelines.
Enterprise Search
Platform
Unified semantic enterprise search over Jira, Confluence, and ServiceNow using Google Cloud Search with custom connectors, metadata enrichment, and relevancy tuning. Delivered 40%+ improvement in knowledge retrieval accuracy.
Largest content and
data migration plan
Architected the 5th largest content and data migration for a primary subsidiary of a leading Hyperscaler. Managed massive-scale data orchestration and integrity across distributed cloud environments.
AI PII Redaction
Engine
Designed an AI-driven solution for a major social media platform (TikTok) to identify and redact PII/sensitive data within massive content streams. Solved a complex data privacy challenge that had remained unresolved by prior vendors for over 6 months. EDUCATION & CERTIFICATIONS
Education
PostGraduate — AI & ML: Business Applications
M.S. Chemical Engineering IISc Bangalore, India
B.S. Chemical Technology Osmania University, India Certifications
Google Professional ML Engineer
Google Professional Data Engineer
Google Looker LookML Developer
Introduction to Quantum Computing — MIT xPRO
Adobe 6.0 Lead Developer Oracle 8i DBA
TECHNOLOGY STACK
GenAI & LLMs Vertex AI Gemini, GPT-4, Claude, Llama, RAG Architectures, Fine-tuned BERT, Prompt Engineering, LLM Evaluation, RLHF Patterns
Agentic AI Vertex AI ADK, LangChain, ReAct, Multi-Agent Systems, Tool-Calling, Function Calling, Memory + Planning Architectures
ML Platforms Vertex AI (Training, Pipelines, Model Registry, Endpoints), Kubeflow, MLflow, CI/CD/CT, A/B Testing, Model Monitoring
Cloud (GCP) Gemini Enterprise, Vertex AI, BigQuery, BigQuery ML, Dataflow, Composer, Dataplex, Cloud Run, Cloud Functions, App Engine, AlloyDB, Spanner, Bigtable Cloud (AWS) Bedrock, SageMaker, S3, Lambda
Cloud (Azure) Azure Databricks
Conversational AI CCAI, Dialogflow CX, Agent Assist, Insights, Custom React Chatbots, Streaming STT/TTS Enterprise Search Vertex AI Search, Google Cloud Search, Retail Search, Custom Connectors (Jira, Confluence, ServiceNow)
Data Engineering Dataflow (Apache Beam), Composer (Airflow), Dataplex, DataStage, ETL/ELT, BigQuery ML, SQL/NoSQL
Analytics & BI Looker, LookML, Tableau, Translation AI, Recommendations AI Languages Python, Java/J2EE, JavaScript/Node.js, PL/SQL, XSLT, DQL
Contact this candidate