0%
Rohit Rathod
AI / ML Engineer — New York
Rohit Rathod

Building production ML systems for financial services and full-stack AI applications that ship. Currently at Fifth Third Bank.

4+
Years in Finance ML
32%
Fraud Detection Lift
4
Projects Shipped
Scroll to explore
ML Engineer
Agentic AI
MLOps
Fraud Detection
LangGraph
Fifth Third Bank
New York
Walk-Forward
ML Engineer
Agentic AI
MLOps
Fraud Detection
LangGraph
Fifth Third Bank
New York
Walk-Forward

About
The full
picture

I'm an AI/ML Engineer at Fifth Third Bank in New York, building production systems across fraud detection, credit risk, and NLP pipelines. My work sits at the intersection of machine learning research and real-world deployment constraints.

Before that I spent three years at Mphasis building financial services ML infrastructure. I hold an MS in Computer Science from Pace University (December 2024) with an obsession for making models that work under load, through data drift, and through audits.

Outside of finance, I build AI-powered consumer products. My focus is always on systems that are genuinely useful, not just technically impressive.

LocationNew York City, NY
CurrentlyAI/ML Engineer, Fifth Third Bank
EducationMS Computer Science, Pace University — Dec 2024
Experience4+ years in financial services ML
FocusMLOps, Agentic AI, Full-Stack AI Products
4+
Years in Finance ML
32%
Fraud Detection Lift
48k+
Predictions/hr
4
Projects Shipped
Work Experience
Work history
Jul 2024 — Present
Fifth Third Bank
AI / ML Engineer
Built real-time fraud detection systems at 48K+ predictions/hr. Improved accuracy by 32% through LightGBM/XGBoost ensemble redesign with walk-forward validation. Architected RAG pipelines for compliance and led MLOps standardization aligned to SR 11-7 model governance.
LightGBMKafkaMLflow PythonLangChainDatabricks SnowflakeSHAPDockerAWS
Nov 2019 — Dec 2022
Mphasis
ML Engineer
Developed credit risk scoring models and NLP pipelines for financial document processing across multiple client engagements. Built low-latency inference services for real-time credit decisions and established model monitoring practices for production health tracking.
XGBoostLSTM TransformersPythonFastAPI PostgreSQLKubernetes
01
Project 01 — Full-Stack AI App
Pawsible

An AI-powered pet adoption and rental platform. Smarter listings, better matches, and a chatbot that actually helps you find the right pet.

Next.js 15Google GeminiGenkitTailwind CSSShadCN UITypeScript
02
Project 02 — MLOps / Trading
AlphaWatch

A self-healing ML trading pipeline that monitors financial instruments for drift and retrains itself automatically. 4,800 lines. 30 files. No human in the loop.

LangGraphLightGBMMLflowAirflowFastAPIKubernetesOptuna
03
Project 03 — Production ML
Fraud Detection

Production fraud detection at Fifth Third Bank. 48K+ predictions per hour on Kafka streams. Sub-20ms inference. 32% accuracy lift over baseline.

KafkaLightGBMSHAPDatabricksAWSMLflow
04
Project 04 — Agentic AI
Fraud Agent

A React application demonstrating a complete agentic fraud investigation workflow. Alert to verdict, with streaming AI, tool-use visualization, and multi-turn case management.

Claude APIReactTypeScriptStreamingTool Use
Technical Skills
The toolkit
ML / Modeling
LightGBM / XGBoostExpert
Feature EngineeringExpert
SHAP / ExplainabilityAdvanced
Drift Detection (PSI/KL)Expert
Optuna HPOAdvanced
LSTM / TransformersAdvanced
Walk-Forward ValidationExpert
MLOps / Infrastructure
MLflowExpert
Apache KafkaAdvanced
Docker / KubernetesAdvanced
Apache AirflowAdvanced
Databricks / SnowflakeAdvanced
AWSProficient
FastAPIExpert
Agentic AI / Full-Stack
LangGraph / LangChainExpert
Claude / Gemini APIsExpert
RAG PipelinesAdvanced
Next.js / ReactAdvanced
Google GenkitAdvanced
TypeScriptProficient
PythonExpert
Contact
Get In Touch
Let's
talk.

Open to senior ML/AI engineering roles in financial services and fintech. Also happy to connect on LangGraph, drift detection, or anything in the agentic AI space.

rohitrathod3272b@gmail.com ↗
Location
New York City, NY
Currently
Fifth Third Bank
Education
MS CS, Pace University
Status
Open to work