Data Engineering
I clean, transform, validate, and prepare large datasets using Python, SQL, feature engineering, and ETL pipelines.
Data Scientist | Machine Learning & Generative AI
I build production-ready machine learning, forecasting, Generative AI, and data-driven systems that turn complex information into practical business and operational insight.
I’m a Data Scientist with five years of experience solving complex business and operational problems with data, machine learning, and AI.
I clean, transform, validate, and prepare large datasets using Python, SQL, feature engineering, and ETL pipelines.
I build, tune, test, and deploy predictive models for real production environments.
I use TensorFlow and PyTorch to model complex patterns that traditional approaches cannot capture effectively.
I build systems that classify, search, extract, and understand information from text.
I developed forecasting systems to estimate vessel behaviour when GPS or satellite signals were limited.
I integrate LLMs into workflows to automate analysis and make complex information easier to use.
I build retrieval pipelines that connect LLMs with relevant documents to produce contextual, grounded responses.
NavSim Technologies
St. John’s, NL, CanadaOwned the end-to-end lifecycle of production ML systems, from experimentation and feature engineering through validation, release, monitoring, and improvement.
Developed forecasting systems to estimate vessel behaviour when GPS or satellite signals were limited, using operational data and robust validation.
Integrated LLMs into internal workflows and built RAG pipelines for context-aware maritime document search and analysis.
Built and maintained FastAPI services so production systems could securely request and consume model predictions.
Created automated accuracy and stability tests plus reusable Python workflows for preparation, preprocessing, and experimentation.
Production-focused machine learning, forecasting, and Generative AI work.
Estimate vessel behaviour when GPS or satellite signals were limited.
Give production applications a dependable way to consume machine learning predictions.
Make internal maritime knowledge easier to find and apply in context.
Identify drift, weak prediction areas, and instability before they affected operational reliability.
St. John’s, Newfoundland, Canada
New Delhi, India
I’m open to roles where I can turn difficult data and AI problems into dependable products people can use.