AI/ML & Generative AI Engineer

Aprotiim Joardar

AI/ML Engineer

Gainesville, FL (open to relocation) 901-***-**** ********.*@************.*** LinkedIn GitHub Streamlit SUMMARY

6+ years of experience in Data Engineering, Machine Learning, and Artificial Intelligence, specializing in Generative AI, LLM Engineering, and scalable cloud-native data platforms, delivering enterprise-grade solutions across research and industry environments. Proven expertise in building RAG architectures, multi-agent AI systems, MLOps pipelines, and large-scale distributed data processing using AWS, PySpark, LangGraph, and modern ML frameworks. Demonstrated success driving measurable business impact through predictive analytics, deep learning, and production AI deployments. SKILLS

Programming Languages: Python, SQL, PySpark, Java, R Artificial Intelligence & Machine Learning: Supervised Learning, Unsupervised Learning, Deep Learning, Neural Networks, NLP, Time Series Forecasting, Feature Engineering, Ensemble Learning, XGBoost, Statistical Modeling, Predictive Analytics Generative AI & LLM Engineering: Retrieval-Augmented Generation (RAG), Agentic AI, Multi-Agent Systems, LangGraph, LangChain, Prompt Engineering, LLM Fine-Tuning, LoRA, PEFT, Vector Databases, Knowledge Graph Integration. Neo4j Machine Learning Frameworks: PyTorch, TensorFlow, Scikit-Learn, Hugging Face Transformers, Keras Data Engineering & Big Data: PySpark, ETL/ELT Pipelines, Data Modeling, Data Warehousing, Distributed Computing, Azure Databricks, Large-Scale Data Processing, Data Quality Management MLOps & Model Lifecycle Management: MLflow, DVC, Experiment Tracking, Model Versioning, CI/CD Pipelines, Docker, FastAPI, Model Monitoring, Automated Retraining, Production Model Serving Cloud & DevOps: AWS (EC2, S3, IAM, SageMaker, Bedrock, Lambda), Kubernetes, Terraform, GitHub Actions Analytics & Visualization: Power BI, Tableau, A/B Testing, Causal Inference, Business Intelligence, KPI Reporting EDUCATION

M.S. in Information Systems & Operations Management (Data Science Concentration) May 2025 University of Florida GPA: 3.91

EXPERIENCE

University of Florida, FL - Data Scientist (Research Assistant) May 2025 – Present

• Developed deep learning models for macroeconomic and financial forecasting using structured economic datasets, improving prediction accuracy and reducing forecast error by 18% compared to baseline forecasting approaches.

• Engineered 25+ predictive features utilizing lag variables, rolling-window statistics, feature interactions, and data transformation techniques to enhance model performance and generalization.

• Optimized neural network architectures through hyperparameter tuning, cross-validation, and regularization techniques, improving model stability and reducing overfitting. ServiceNow, FL - AI/ML Engineer Nov 2024 – May 2025

• Architected and deployed enterprise-grade Retrieval-Augmented Generation (RAG) platforms leveraging LangGraph, AWS Bedrock, vector databases, and semantic search frameworks, improving knowledge retrieval accuracy by 42%.

• Engineered multi-agent AI workflows utilizing LLM orchestration, autonomous task routing, contextual memory management, and FastAPI services, reducing incident resolution time by 48% across more than 500 microservices.

• Built and maintained Neo4j knowledge graphs integrating ServiceNow enterprise data to provide contextual retrieval for AI- powered workflows, improving incident resolution accuracy and operational insights.

• Implemented cloud-native AI infrastructure on AWS utilizing SageMaker, Bedrock, EC2, IAM, and S3 services, enabling secure, scalable, and highly available machine learning workloads. KPMG, India - Senior Data Engineer Jul 2018 - Jul 2023

• Led the architecture and implementation of enterprise-scale data engineering solutions supporting finance, HR, and operational analytics initiatives across SAP, Oracle, and Workday ecosystems, processing more than 20TB of structured.

• Designed highly scalable PySpark-based ETL and ELT pipelines utilizing distributed computing principles, significantly improving data processing throughput, pipeline resiliency, and downstream analytical performance.

• Developed cloud-integrated data platforms supporting advanced analytics, machine learning initiatives, and executive reporting requirements across multiple business functions and stakeholder groups.

• Optimized large-scale data transformation workflows through partitioning strategies, query tuning, caching mechanisms, and distributed processing techniques, reducing processing latency and improving platform efficiency.

• Mentored and managed a team of data engineering professionals, establishing development standards, code review practices, and delivery processes that improved project quality and execution consistency. PROJECTS

CarGenuity – AI - Powered Used Car Buying Assistant – GitHub Link

• Built CarGenuity, an AI-powered used car buying assistant, using a Next.js 14 frontend and FastAPI backend with a LangGraph multi-agent orchestration pipeline across search, RAG, and analysis agents using Claude models.

• Designed a RAG retrieval pipeline backed by Pinecone vector search and S3 storage, surfacing relevant buying-guide content in real time and reducing first-time buyer decision friction through context-aware AI responses. Agentic AI Blog System GitHub-Link

• Designed a multi-agent AI system supporting task planning and parallel execution, enabling automated content generation workflows with improved efficiency

• Reduced content generation latency by 35% by implementing parallel agent execution and optimizing orchestration logic YouTube Comments Sentiment Analysis GitHub-Link

• Built and deployed an NLP-based sentiment analysis system as a Chrome extension on AWS using Code Deploy

• Improved model accuracy from 60% to 87% through model selection, feature engineering, and hyperparameter tuning Corrective RAG Knowledge System GitHub-Link

• Built a corrective RAG pipeline to reduce hallucinations in retrieval-based systems, improving response reliability

• Reduced hallucination rates by 40% using retrieval confidence scoring and adaptive web augmentation Production Chatbot Deployment GitHub-Link

• Deployed a production-grade chatbot with RAG, HITL and MCP server integration for scalable interaction handling

• Improved resolution accuracy by 30% through caching strategies, fallback mechanisms, and prompt optimization

Contact this candidate