Welcome -

Aniket Shirsat | Data Scientist

Featured AI & Data Projects

Current focus: enterprise multi-agent workflows, SQL Server + OCR document intelligence, production RAG systems, and cloud AI automation for business operations.

? Multi-Agent Insurance AI Platform

Enterprise GenAI and Agentic Workflow System

Business Problem: Insurance and operations teams spend significant time extracting, validating, and reviewing unstructured documents manually
Solution: Built a modular multi-agent workflow for document extraction, reasoning, validation, and structured outputs using LLM orchestration
Recruiter Signal: Demonstrates applied GenAI, agentic systems, enterprise workflow automation, and production-oriented system thinking

View Code LangGraph OpenAI Agentic AI Python

?? Vessel Fuel Optimization Engine

Enterprise Machine Learning Solution

$1M Annual Savings: Deployed XGBoost and time-series analysis model achieving 5% fuel consumption reduction across 50+ vessels
Advanced ML Pipeline: Built end-to-end predictive analytics system with real-time optimization algorithms
Maritime Industry Impact: Improved fuel efficiency for commercial shipping operations with data-driven insights

Enterprise ML XGBoost Time Series

?? AI-Powered PDF Knowledge Assistant

Intelligent Document Processing

70% Ticket Reduction: Built AI chatbot using OpenAI and LlamaIndex for instant PDF document queries
24/7 Automation: Implemented RAG (Retrieval-Augmented Generation) architecture for context-aware responses
User Experience: Enhanced document search and support workflows with intelligent document parsing and vector-based retrieval

View Code OpenAI LlamaIndex RAG

? Automated Invoice Processing Platform

AWS Cloud Automation Platform

90% Time Reduction: Built document processing automation pipeline using AWS Textract, S3, and SES
Enterprise Integration: Developed AI-powered document processing with real-time notifications
Scalable Architecture: Created end-to-end pipeline from document capture to automated expense management

View Code AWS Textract OCR

??? AWS Computer Vision & Content Safety Suite

Multi-Service AI Platform

Content Moderation: Automated inappropriate content detection with confidence scoring using AWS Rekognition
Text Intelligence: Smart text extraction from documents, receipts, and forms with Amazon Rekognition
Platform Safety: Enhanced user-generated content experiences with real-time filtering and analysis

Moderation Text Detection Computer Vision

?? AWS Cloud Automation Toolkit

Enterprise Cloud Infrastructure

Storage Automation: Built comprehensive S3 operations suite for scalable cloud storage management
Email Infrastructure: Engineered mass email automation using AWS SES for marketing campaigns
Cost Optimization: Implemented intelligent data archiving and lifecycle management for enterprise applications

S3 Suite SES Platform AWS Infrastructure

?? Machine Learning Prediction Systems

Applied ML and Research Projects

Loan Prediction: 82% accuracy system using ensemble methods (Random Forest, XGBoost) with 40% time reduction
Sentiment Analysis: Real-time Twitter analysis with interactive Streamlit dashboard and 95% user satisfaction
Cultural AI Research: 90% accuracy model for detecting cultural ambiguity in images using advanced NLP techniques

Machine Learning Streamlit Computer Vision

?? Impact Summary

$1M+

Annual Savings Generated

70%

Process Automation

50+

Vessels Optimized

90%

AI Model Accuracy

10+

End-to-End AI Projects

Professional Experience

Data Scientist

Nautarex Systems Pte Ltd

Jan 2026 – Present

Designed and developed an AI-powered maritime document intelligence platform using Google Vertex AI Gemini to automate extraction of critical data from Bunker Delivery Notes (BDNs) received in diverse formats and languages.
Built an end-to-end enterprise automation pipeline integrating Python, SQL Server, Google Vertex AI, OCR technologies, and business-rule validation to process, enrich, and persist maritime fuel delivery data.
Engineered domain-specific prompt frameworks and validation rules to improve extraction accuracy for fuel quantity, sulfur content, fuel grade, port information, and regulatory compliance across global marine operations.
Architected a scalable, production-grade document AI framework with automated exception handling, logging, and data quality checks to streamline maritime compliance workflows and reduce manual processing effort.

Research Assistant (NLP & LLM)

Indiana University Indianapolis

Jan 2025 – May 2025

Crafted a dataset pipeline for analyzing annotated advertisement images, enhancing data quality by 20%, and improving research efficiency through vision-language models and automated testing frameworks
Engineered a large language model to detect cultural ambiguity in images, achieving 90% accuracy in identifying cultural nuances, by leveraging advanced NLP techniques and cross-validation methods

Data Scientist

Nautarex Systems Pte Ltd

Jun 2023 – Aug 2024

Automated invoice importing and two-way matching using Amazon Textract, AWS S3, and AWS SES, reducing manual processing time by 90%, and enhancing operational efficiency through cloud-based automation
Developed a corporate reimbursement app using Flutter and AWS services, improving employee expense management efficiency by 25%, through integration of AI-powered document processing and real-time notifications
Deployed a vessel fuel optimization model using XGBoost and time-series analysis, achieving a 5% reduction in fuel consumption across 50+ vessels, resulting in $1 million annual savings through strategic data-driven insights
Built a customer churn prediction model using Random Forest algorithms, achieving 85% accuracy and reducing customer churn by 5% through targeted retention campaigns and data-driven decision-making

Data Analyst

Capgemini

Feb 2022 – May 2023

Developed 5 data analytics solutions and 8 BI dashboards in Python and Tableau, reducing reporting time by 30%
Optimized 20+ SQL queries across MS SQL and Oracle databases, improving query performance by 40%
Led data migration for 3 business systems with 99.9% accuracy, reducing validation time by 50%
Created 12 automated visualization reports for financial clients, achieving $200K annual cost savings

Supply Chain and Logistics Intern

Frost & Sullivan APAC

Jan 2021 – May 2021

Analyzed sales data to improve client product lines, resulting in a 15% increase in overall sales through proposed pricing and sales strategies, by conducting market research and competitor analysis
Conducted industry research on emerging technologies, enhancing market intelligence by 30%, through analyzing revenue, investments, sales, and pricing strategies

Business Development Executive

Nautarex Systems Pte Ltd

Mar 2019 – Aug 2020

Conducted market research and prepared value propositions for product cross-selling, reducing customer churn from 5.1% to 3.9%, by building strong customer relationships and implementing targeted marketing strategies
Led new prospect pitching and sales processes, achieving an 11% revenue increase within 10 months, through strategic sales planning and effective client engagement
Streamlined customer post-sales support processes, reducing overall costs by 7%, by implementing efficient communication channels and proactive issue resolution

Infantry Trooper / Finance Clerk & Assistant

Singapore Armed Forces

Feb 2017 – Feb 2019

Led a team of 5 colleagues in executing a successful mock anti-terrorist operation, enhancing team coordination and tactical efficiency by 70%, through strategic planning and effective communication
Managed financial operations and administrative tasks with precision and accountability in high-pressure military environment

Technical Skills

AI & LLM Engineering

Google Gemini (Vertex AI) Google Vertex AI OpenAI API LangChain LangGraph LlamaIndex Ollama Retrieval-Augmented Generation (RAG) Prompt Engineering OCR Vector Search (FAISS) ChromaDB Agentic AI

Machine Learning

Scikit-learn TensorFlow PyTorch Hugging Face XGBoost Deep Learning Reinforcement Learning

Cloud & Infrastructure

Google Cloud Vertex AI GCP Document AI AWS S3 AWS Textract Amazon SageMaker Amazon Bedrock AWS Lambda AWS SES Amazon Rekognition Amazon Comprehend Azure Databricks Azure Data Factory Databricks Lakehouse

Software Engineering

Python R SQL Flutter FastAPI REST APIs Git GitHub Agile/Scrum

Analytics & Visualization

Pandas NumPy Statistical Analysis Hypothesis Testing Tableau Power BI Plotly Matplotlib

Professional Certifications & Continuous Learning

IBM RAG & Agentic AI IBM Generative AI Engineering AWS Generative AI Immersion Day Stanford Machine Learning IBM Deep Learning & Reinforcement Learning IBM Supervised Machine Learning IBM Unsupervised Machine Learning Google Cloud Enterprise Database Migration Databricks Lakehouse Fundamentals Dataiku Developer Dataiku Advanced Designer Dataiku ML Practitioner MITx Python Java Programming Agile Software Development BigID Professional OneTrust Privacy Professional OneTrust GRC Professional OneTrust Consent & Preference Management OneTrust Vendor Risk Management

Aniket Shirsat

Featured AI & Data Projects

? Multi-Agent Insurance AI Platform

?? Vessel Fuel Optimization Engine

?? AI-Powered PDF Knowledge Assistant

? Automated Invoice Processing Platform

??? AWS Computer Vision & Content Safety Suite

?? AWS Cloud Automation Toolkit

?? Machine Learning Prediction Systems

?? Impact Summary

Professional Experience

Technical Skills

AI & LLM Engineering

Machine Learning

Cloud & Infrastructure

Software Engineering

Analytics & Visualization

Professional Certifications & Continuous Learning

Education

About Me

Get In Touch

Aniket’s AI Assistant