My Projects

A showcase of large-scale enterprise data architectures, applied machine learning systems, and independent open-source data science tools.

Professional Projects

Note: The projects listed below were architected and executed for enterprise clients. Specific client names, sensitive data, and proprietary business logic have been generalized or omitted to comply with confidentiality agreements.

  • AI-Driven Marketing Copy Generation System
    Led the architecture and deployment of an enterprise data application leveraging Snowflake Cortex to dynamically author marketing and SEO-optimized product descriptions at scale across multiple international regions.
    Snowflake Cortex Prompt Engineering Snowpark Python
  • Intelligent Sales Agent for Analysts
    Developed cloud-native LLM-powered agents to assist analysts in querying complex sales and planning data via natural language. Delivered on-demand insights directly to stakeholders, significantly reducing report sprawl and manual dashboard creation.
    Snowflake Cortex LLM Agents RAG Architecture Semantic Search
  • Intelligent Data Model Mapping Application
    Led the development of a native cloud data application to dynamically map data from diverse models to standard enterprise schemas. Built automated ELT workflows for data transformation, filtering, and mapping. Engineered version control, CI/CD, and leveraged integrated LLMs via Snowflake Cortex for in-app user guidance.
    Snowflake Native Apps Snowflake Cortex dbt (Data Build Tool) GitHub Actions
  • Automated PII Detection and Tokenization Native App
    Built a Snowflake Native App for sensitive data detection to meet strict compliance requirements. Orchestrated Transformer models for known entity detection alongside Cortex LLMs for unstructured data. Enabled in-transit tokenization via external API integration using the Very Good Security (VGS) vault, seamlessly extending into dynamic data masking policies.
    Snowflake Native Apps External Functions Transformer Models Dynamic Data Masking
  • Petabyte-Scale NLP Monitoring Platform
    Engineered a state-of-the-art multilingual monitoring platform processing over 1 billion documents and communication records. Integrated advanced semantic search, entity extraction pipelines, and knowledge graph clustering for robust 'needle in the haystack' event tracing deployed on distributed cloud infrastructure.
    Named Entity Recognition (NER) Semantic Search Knowledge Graphs AWS EC2 / S3
  • Cloud-Native Centralized Feature Store
    Developed a highly scalable, centralized feature store utilizing modern cloud-native capabilities. Designed the architecture to support feature reuse, automated online synchronization, and robust role-based access control (RBAC) protocols across disparate data science teams.
    Coalesce Snowflake Role-Based Access Control (RBAC)
  • Deep Learning Pathogen Identification System
    Developed a multi-pathogen detection system employing complex deep learning architectures. Utilized both supervised and unsupervised methods in low-data regimes, ensuring rigorous model testing and validation across cloud-hosted machine learning environments.
    AWS SageMaker PyTorch Convolutional Neural Networks (CNN) Transfer Learning
  • Portfolio Analysis and Cashflow Prediction Engine
    Designed high-throughput data ingestion architectures from disparate financial providers into a centralized cloud warehouse. Implemented time-series forecasting models directly within the data layer to predict cashflows, visualized via comprehensive business intelligence dashboards.
    Snowflake Snowpark Time-Series Forecasting ARIMA / Prophet Sigma
  • Scalable Demand Forecasting Engine
    Implemented extensive machine learning pipelines to forecast product demand across varying time horizons. Enabled optimized raw material procurement workflows by deploying distributed processing models and streaming predictions to executive reporting interfaces.
    Databricks Spark SQL XGBoost / LightGBM PowerBI

Independent Projects

View all on GitHub
  • PCvalueBuilder: A recommender system to assist customers in upgrading their PC hardware
  • A convolution neural network model for Spoken digit recognition using Mel-frequency Cepstrum
  • Implementation of image classifier for RGB images using artificial neural networks (NN)
  • Optimal image compression for efficient reconstruction using unsupervised learning
  • Air quality prediction using hyperparameter regression analysis
  • Design and Analysis of a Feed-forward PUF (32 nm node) under Voltage Scaling
  • Nanoscale Implementation of G-Share Branch Predictor
  • An AES-Based Intellectual-Property Identification in System-on-a-Chip (SOC) Design