Open to opportunities

Aryan Agarwal

@aryanagarwal

Message

AI/ML Associate Consultant building enterprise NLP, RAG, and agentic systems.

Zimbabwe

Message

What I'm looking for

I’m looking for a team where I can architect agentic AI/NLP and RAG systems in production, work closely with engineering and stakeholders, and keep improving model reliability, evaluation, and deployment across GCP/AWS. I value impact, clarity, and fast iteration.

I’m an AI/ML Associate Consultant who builds production-grade NLP and GenAI solutions that automate real business workflows. At KPMG, I lead AI architecture for an enterprise-grade NLP grievance chatbot system on GCP Vertex AI.

I implemented custom intent classification and entity extraction that automate 20% of manual routing. I also developed an automated Vertex AI pipeline to train and deploy a high-granularity classifier for a 60-class dataset of 40k+ records, improving feature preparation and model selection.

To push capability further, I fine-tuned Gemini 1.5 Flash using SFT on a 60k-sample NLP dataset for a two-tier grievance classification architecture—boosting categorization granularity and downstream routing precision. I also engineered a proof-of-concept Agentic AI system with LangGraph, designing multi-agent workflows that use MCP-based tool calling for deterministic financial decision-making.

Earlier, as a Data Science Graduate Assistant, I developed a Retrieval-Augmented Generation (RAG) pipeline for TitanGPT, integrating internal research APIs to provide real-time external context and standardize student FAQ resolution. I built evaluation-backed quantitative models on Vertex AI to identify default-risk drivers, fine-tuned FinBERT for sentiment-driven signals from 10-K filings, and automated XGBoost training on AWS SageMaker to improve reproducibility and reduce setup time.

Experience

Work history, roles, and key accomplishments

Current

Associate Consultant (AI/ML)

Current

KPMG

Jul 2025 - Present (11 months)

Led AI architecture for an enterprise NLP grievance chatbot on GCP Vertex AI for PFRDA, implementing intent classification and entity extraction to automate 20% of manual routing. Built and deployed a Vertex AI NLP pipeline for a 60-class, 40k+ record dataset and fine-tuned Gemini 1.5 Flash for two-tier grievance classification to improve categorization granularity and routing precision.

Vertex AI NLP LangGraph Multi Agent Systems

Data Science Graduate Assistant

California State University, Fullerton

Aug 2024 - Jun 2025 (10 months)

Developed a Retrieval-Augmented Generation (RAG) pipeline for TitanGPT using internal research APIs to provide real-time external context and standardize student FAQ resolution. Designed LangGraph-based RAG workflows, built default-risk models on Vertex AI, fine-tuned FinBERT for 10-K sentiment-driven risk signals, and automated XGBoost training on AWS SageMaker to improve reproducibility for stud

Retrieval Augmented Generation (RAG)LangGraph Vertex AI FinBERT AWS Sagemaker XGBoost Financial NLP

Analyst (Data)

TheEsportsClub

Oct 2020 - May 2021 (7 months)

Performed customer segmentation by applying k-means clustering on 100K+ survey responses, improving ad targeting efficiency by 30%. Ran A/B testing with statistical hypothesis tests to quantify user engagement impact, reducing churn risk by 20%, and built interactive Tableau dashboards for competitive player statistics.

Customer Segmentation A B Testing Hypothesis Testing Tableau Data Visualization Survey Analytics

Education

Degrees, certifications, and relevant coursework

California State University, Fullerton

Master of Science (M.S.), Information Systems

2023 - 2025

Grade: 3.86/4.0

Activities and societies: Graduate Assistant (two departments); featured in CSUF newsletter for helping develop and deploy TitanGPT.

Earned an M.S. in Information Systems from California State University, Fullerton, graduating with a 3.86/4.0 GPA. Served as a Graduate Assistant and contributed to helping develop and deploy TitanGPT.