PORTFOLIO / ANALYTICS · ML · AI

Work that turns analysis
into decisions.

A curated body of applied analytics, machine learning, and decision-support projects. Each case study is organized around the problem, evidence, outcome, and practical use—not just the technology used to build it.

14Public projects
6Focus areas
5Live applications
Project index

Browse by expertise

A focused index of analytics, machine learning, AI, and decision-support work.

Showing 14 of 14 public projects

Latest projectEnergy Analytics · Forecasting · Live Operations

GridPulse Energy Grid Analytics

An operational energy analytics system built on EIA-930 PJM hourly grid data, combining forecast benchmarking, source-data QA, stress screening, error intelligence, and a live/replay Streamlit control room.

44.2%Lower overall MAE
View case study
02Anomaly Detection · Environment

Water Quality Analysis

A project that searched water measurements for unusual patterns so potential pollution, sensor problems, or emerging environmental risks could be investigated sooner.

3Analytical methods
View case study
03Classification · Healthcare

Healthcare Resource Modeling

A comparison of machine learning methods designed to help planners understand which cases may require resources and how confidently a model can support that decision.

3Classifiers compared
View case study
04NLP · Deployment

Sentiment Analyzer

A plain language tool that reads a written review and estimates whether the overall opinion is positive or negative, helping large volumes of feedback become easier to summarize.

91.19%Test accuracy
View case study
05RAG · Applications

Sports Q&A Chatbot

A question answering assistant that looks through a focused sports knowledge base before responding, making answers more grounded than relying on a language model's memory alone.

10Source documents
View case study
06LLM Fine tuning · Research

Medical Q&A Model

A research prototype that taught a large language model to better follow medical question and answer patterns without retraining all seven billion model parameters.

0.6573Final validation loss
View case study
07Computer Vision · Explainability

Skin Disease Classifier

An explainable image classification prototype that suggests a skin lesion category and highlights the image regions that most influenced the prediction.

74.15%Test accuracy
View case study
08Computer Vision · Zero Shot Detection

Open Vocabulary Object Detector

An interactive vision tool that lets a person describe an object in ordinary words and then asks the model to locate matching items in an uploaded image.

80 → openVocabulary expanded
View case study
09Computer Vision · Medical Imaging

Skin Lesion Segmentation

A medical imaging prototype that traces the pixel level boundary of a skin lesion, helping show its shape and extent rather than returning only a category name.

0.9115Test Dice score
View case study
10LLM Evaluation · Medical AI

LLM Evaluation Framework

A framework that compared a general Mistral 7B model with its medically fine tuned version to test whether specialization actually improved the answers.

+17.8%Average ROUGE gain
View case study
11Environmental Analytics · Decision Intelligence

PFAS Drinking Water Decision Intelligence

An end-to-end drinking-water decision-support project that turns nearly two million EPA monitoring records into an auditable engineering-review queue, utility case studies, lifecycle economics, and constrained infrastructure funding portfolios.

1.99MAnalytical-result records
View case study
12Infrastructure Analytics · Asset Management

Wastewater Infrastructure Analytics

A reproducible wastewater asset-management and capital-planning framework that connects utility data ingestion, SQL quality assurance, transparent risk scoring, lifecycle cost analysis, and constrained investment prioritization.

6Decision stages
View case study
13Financial Crime Analytics · Graph ML

Financial Crime Risk Intelligence

A large-scale Elliptic2 AML decision-support project combining out-of-core graph feature engineering, validated model selection, constrained-review prioritization, calibration, and case-specific statistical evidence.

0.528Repeated PR-AUC
View case study
14Global Analytics · Interactive Dashboard

World Happiness Dashboard

An animated 2015–2019 global dashboard for country rankings, regional patterns, factor relationships, and an out-of-time 2019 regression evaluation—running natively inside the portfolio.

782Country-year observations
Open dashboard