Data Science & Machine Learning

Data Analytics & Visualization

Data Engineering & Infrastructure

Cloud & Full Stack

Development Tools & Automation

Sustainability & Environmental

🗓️ Oct 2025 – Present   |   📍 Tempe, AZ

• Built a production IoT platform for 9 stations, 14 sensor parameters, and 1.58M+ records.
• Achieved 99.9997% ingestion success with idempotent batching and automated reconciliation.
• Tracked 132 MLflow runs across 5 anomaly-model families, reaching 0.80 F1 and 0.85 recall.
• Cut fleet-wide retrieval from hours to under 2 seconds; delivered the platform for active SRP use.

Skills: Python · Machine Learning · MLflow · FastAPI · TimescaleDB · AWS

🗓️ Feb 2025 – Oct 2025   |   📍 Tempe, AZ

• Automated a federal-data ELT pipeline, reducing processing time by 60%.
• Optimized traversal of a 1.07M-node river graph from 95 seconds to 12 seconds.
• Modeled flow for 337K ungauged river reaches with Gradient Boosting, achieving 0.87 R².
• Engineered geospatial features used by 3 research groups for infrastructure-risk analysis.

Skills: Python · scikit-learn · ArcGIS Pro · QGIS · ETL · Graph Algorithms

🗓️ May 2024 – Dec 2024   |   📍 India

• Delivered production AI and backend features for a 5,000+ user learning platform.
• Improved retention prediction by 18% using logistic regression and XGBoost on 500K+ records.
• Deployed a LangChain and OpenAI RAG assistant that reduced support tickets by about 30%.
• Cut PostgreSQL p99 latency by 40% and ingestion errors by 80% on AWS.

Skills: Python · FastAPI · XGBoost · LangChain · RAG · PostgreSQL · AWS

🗓️ May 2023 – Aug 2023   |   📍 India

• Analyzed 25K+ campaign interactions across 3 channels using SQL and Python.
• Reduced weekly reporting from 3–4 hours to about 1 hour with automated data pipelines.
• Built 2 Tableau dashboards tracking 8 acquisition, funnel, and experimentation KPIs.
• Measured a 15% relative CTR lift through an A/B test involving 3K+ visitors.

Skills: SQL · Python · dbt · Tableau · Excel · A/B Testing

AI-native food-truck marketplace spanning 107 trucks and 635 menu items. A 4-tier stack (Next.js, Spring Boot gateway, FastAPI/LangGraph AI service) powers an AI Concierge, a structured-output Order Builder, and an Owner Copilot for sales and complaint analysis.

Focus: Agentic AI · Full-Stack AI Product · Tool-Calling
Tech: Next.js · Spring Boot · FastAPI · LangChain/LangGraph

Agentic medical RAG platform indexing 9K+ medical chunks with FAISS and cross-encoder reranking, plus a 3-step ReAct trace for auditable answers. Benchmarked across 100 prompts and 5 medical domains for grounded accuracy.

Focus: Medical RAG · Agentic AI · Healthcare
Tech: ChromaDB · FAISS · LangChain · Next.js

Fashion computer-vision and hybrid recommendation engine combining CLIP embeddings with cold-start-aware weighting. Reaches 0.73 Precision@10 and 89% catalog coverage, with GPT-4o-mini explaining every recommendation.

Focus: Computer Vision · Recommender Systems
Tech: PyTorch · CLIP · FastAPI · Next.js

AIOps anomaly detector polling Kubernetes pod memory every second, using an LSTM + LightGBM hybrid to forecast failures before they happen. A custom Go Prometheus exporter feeds predictions into the autoscaler, with LLM-generated remediation commands.

Focus: AIOps · Kubernetes · SRE
Tech: TensorFlow · LightGBM · Go · Prometheus

End-to-end dissolved-oxygen forecasting and sensor-fault triage across 14 USGS stream sites in 11 states. Per-site gradient-boosting models reduced forecast error by 25% at 6 hours, while a cost-aware alert layer warned of 80% of low-oxygen events at very low alert volume.

Focus: Environmental ML · Time-Series Forecasting · Sensor Analytics
Tech: Python · scikit-learn · FastAPI · Docker · GitHub Actions

Retail demand forecasting across 1,115 stores. BigQuery ML ARIMA_PLUS models generate 33,450 30-day forecasts, benchmarked at 16.92 MAPE versus 21.91–24.74 for baseline models, delivered through a FastAPI backend and Streamlit dashboard.

Focus: Time-Series Forecasting · Data Science
Tech: BigQuery ML · FastAPI · Streamlit · Python

Deep Learning Specialization
DeepLearning.AI

Optimization for real-world AI systems