Arzoo Jiwani

Applied AI / Machine Learning Engineer

I enjoy turning messy, real-world problems into reliable machine learning systems from search and recommendations to production-ready AI services.

Graduating April 2026 · Actively seeking full-time AI / ML roles

Experience building production AI systems in healthcare, search, and recommendation domains.

GitHubLinkedInResume

About Me

I'm an MS in Artificial Intelligence student at Northeastern University, Boston, graduating April 2026. I build AI systems that solve real problems and drive measurable business impact from smart recommendations to production-ready AI services.

My experience spans retrieval-augmented generation (RAG), recommendation engines, and AI-driven microservices. I focus on creating end-to-end AI pipelines that are reliable, scalable, and integrated into workflows, ensuring solutions not only work technically but deliver value where it matters.

I thrive at the intersection of data, models, and business goals, turning messy, real-world problems into actionable insights and AI solutions that are fast, explainable, and maintainable. Whether it’s healthcare, search, or recommendations, my aim is to make AI both impactful and understandable for decision-makers and end users.

Featured Projects

RAG-based AI Search Bot for University Research Portal

  • • Designed and developed a Retrieval-Augmented Generation (RAG) search engine for semantic research discovery
  • • Implemented RoBERTa-based query expansion and embedding generation
  • • Built FAISS-based vector retrieval and ranking pipelines
  • • Automated the full NLP workflow, improving search result precision by ~40%
PythonNLPRoBERTaFAISSVector EmbeddingsLLMs
View on GitHub

Spotify-Based Concert & Airbnb Recommendation System

  • • Built a personalized recommendation pipeline using Spotify listening data
  • • Integrated Spotify and Ticketmaster APIs for live event recommendations
  • • Developed an XGBoost regression model for Airbnb recommendations
  • • Achieved an ~85% improvement in prediction accuracy through feature engineering
PythonXGBoostscikit-learnAPIsPandas
View on GitHub

Federated Learning & Blockchain

  • • Co-authored a research project combining federated learning and blockchain
  • • Focused on privacy-preserving model training and decentralized trust
Federated LearningBlockchainPython

Experience

Software Development Intern

July 2025 – December 2025

Darby

  • • Led the development of the core AI microservices backbone for a Medicare coverage determination platform, architecting scalable Node.js services integrated with a Spring Boot backend to support long-running, production AI workflows
  • • Owned the design and iteration of LLM-driven prompt frameworks and decision logic, leveraging FAISS-based vector retrieval to identify relevant NCD/LCD policies and achieving 90%+ accuracy through evaluation on real clinical documents
  • • Built and deployed a production-grade AI microservice for Medicare policy identification, processing unstructured clinical text and structured JSON inference to reduce end-to-end latency to 30–60 seconds
  • • Automated data auditing and quality monitoring pipelines using Python, Pandas, and SQL, fully replacing manual audits with scalable validation, metrics, and visualization workflows
  • • Collaborated cross-functionally with Product and Engineering teams to define data workflows, validation logic, and API contracts, translating evolving business requirements into maintainable AI-driven system design
PythonNode.jsJavaSpring BootSQLDockerREST APIsClaude AIFAISSGit

Skills

Languages

PythonJavaC++SQL

AI / ML

Machine LearningApplied AILLMsRAGRecommendation SystemsNLPXGBoostModel EvaluationFeature Engineering

Data

PandasNumPyFeature EngineeringModel Evaluation

Systems & Tools

APIsDockerGitFAISSBackend Integration

Contact

Open to full-time Applied AI / Machine Learning roles starting April 2026.

EmailGitHubLinkedIn