Open to opportunities

Rohan Swain

@rohanswain

Message

AI engineer specializing in backend systems and LLM-driven analytics.

United States

Message

What I'm looking for

I seek roles building production-ready AI/analytics platforms where I can integrate LLMs, backend systems, and ML models to deliver explainable, reliable insights for cross-functional teams.

I am an AI engineer with strong backend expertise building AI-powered analytics systems across sports and financial services. I design and deploy LLM-driven query interfaces, predictive modeling workflows, and production-ready APIs that enable non-technical stakeholders to access actionable insights.

My recent work includes implementing LangChain-based schema-aware SQL generation, FastAPI backends, and automated ETL pipelines for GPS tracking and event data, reducing ad-hoc analysis time from hours to minutes. I integrate machine learning models (Random Forest, XGBoost) with interpretability techniques like SHAP and maintain rigorous evaluation harnesses for LLM outputs and SQL generation.

I also have experience as a business systems analyst translating regulatory and operational requirements into technical specifications, owning API integrations (MuleSoft), and automating validation and reconciliation workflows to improve release reliability and data integrity. I prioritize reliable, secure, and explainable analytics for cross-functional stakeholders.

Experience

Work history, roles, and key accomplishments

Sports Analyst Intern

University of Delaware

Aug 2025 - Dec 2025 (4 months)

Built a LangChain-based LLM analytics service translating natural language into schema-aware SQL and FastAPI backends, reducing ad-hoc analysis time from ~1–2 hours to under 2 minutes and integrating predictive models (ROC-AUC 0.78) across 50,000+ shot events.

Lang Chain fastAPI SQL Random Forest XGBoost SHAP ETL Faiss

Business Systems Analyst

Concentrix

Jul 2021 - Dec 2023 (2 years 5 months)

Served as systems liaison for loan protection insurance products, authored technical specifications, prioritized JIRA backlogs to reduce sprint spillover by 16%, and designed MuleSoft integrations that cut post-release integration defects by 20%.

MuleSoft Python API Integration Salesforce Jira Data Validation ETL Root Cause Analysis

AI Engineer (Project)

CrimeScope

Designed an agent-based GenAI crime analytics platform combining RAG (FAISS) and forecasting, built a Flask backend with local LLM inference, and implemented automated RAG evaluation to validate answer faithfulness across 10,000+ records.

Faiss Flask Ollama Ragas Embeddings Python

ML Engineer (Project)

Song Energy Classification

Developed and tuned classification models on 30,000+ songs, applying feature scaling and GridSearchCV to achieve 81% accuracy with KNN and identify key features like loudness and acousticness.

scikit learn K NN SVM Random Forest GridSearchCV Cross Validation Python

Education

Degrees, certifications, and relevant coursework

University of Delaware

Master of Science, Business Analytics and Information Management

Grade: 3.93 GPA

Master of Science in Business Analytics and Information Management with a 3.93 GPA, focused on analytics, data integration, and deploying ML-driven solutions.