AI integration + backend
AI Sales CRM Assistant
Structured lead analysis, deterministic urgency normalization and duplicate-safe HubSpot / Google Sheets synchronization.
Open case studyWORK / PROJECT DIRECTORY
Systems, from idea to evidence.
I build and evaluate AI/ML systems with a focus on how models behave in real workflows, how reliably they generalize, and how well the surrounding software holds up. Explore the projects to see the decisions, trade-offs, and evidence behind the work.
9 projects / all capabilities
Hybrid ML + RAG
Intelligent Patient Admission Triage Assistant
Deterministic ESI/MEWS scoring and ML risk prediction, separated from retrieval-grounded explanations.
Open case studyAgentic AI + retrieval
AutoCare Agent
A six-node LangGraph diagnostic workflow connecting hybrid retrieval, four tools and grounded decision support.
Open case studyBiomedical AI research
Swin Transformer for EEG Seizure Classification
Validated 19-channel EEG → FRP preprocessing, completed four-class Swin segment experiments and a separate three-class patient pilot; CNN-vs-Swin comparison remains ongoing.
Open case studyRetrieval engineering
GenAI Research Assistant
A local research Q&A application with reproducible FAISS Flat, HNSW and Annoy retrieval benchmarks.
Open case studyComputer vision + MLOps
Poultry Disease Classification
Four-class VGG16 transfer learning with DVC, FastAPI inference and Docker / GitHub Actions / AWS ECR + EC2 deployment.
Open case studyBiomedical research
EEG Signal Processing & Classification: Sleep Spindle Detection and Motor Imagery Classification
Two reproducible EEG pipelines for sleep spindle detection and left-vs-right motor imagery classification, with interpretable signal processing and leakage-aware session evaluation.
Open case studyAgentic workflows
Agentic LLM Evaluation Workflow for Resume Screening
A LangGraph workflow evaluates resume–job fit through typed extraction, gap analysis, deterministic score calibration and recommendation generation, with LangSmith tracing.
Open case studyData Engineering
Distributed Data Processing & Scalability Analysis
Reproducible PySpark ETL and single-node versus multi-node Spark benchmarking on NYC Taxi data.
Open case study