Machine Learning Consultant2018 - 2018Hack Partners
Technologies: Plotly, Dash, Pandas, NLTK, Python
- Consulted on data valuing for a UK government's civil service organization, the Office of Rail & Road; advised on how to best leverage their data and viable use cases of automated decision making and machine learning.
- Built POC machine -earning/probabilistic models with an interface through an interactive dashboard.
- Ensured the deployment of the entire system easy for the organization's employees.
- Implemented natural language processing on vast numbers of daily logs and made the insights accessible to employees through an interactive dashboard.
- Consulted with senior executives on project scope, timelines, and expected outcomes.
Data Scientist2017 - 2018CarTrawler
Technologies: Amazon Web Services (AWS), AWS, Redshift, Unit, Plotly, Dash, Flask, Pandas, Python, Learning
- Researched about state-of-the-art reinforcement learning techniques for pricing optimization.
- Automated decision-making directly affecting the top and bottom-line performance of the company.
- Built a platform for the orchestration, reporting, and maintenance of online learning pricing algorithms.
- Developed an extensive and flexible interactive dashboard for a pricing optimization platform used by business members outside of data science.
- Worked as part of the revenue management team.
- Communicated with business stakeholders.
Senior Data Engineer2016 - 2017Blue Vision Labs
Technologies: Amazon Web Services (AWS), PostGIS, PostgreSQL, AWS, Python
- Deployed and maintained large-scale data collection operations across the UK and US; this included the hiring, onboarding, and management of data collection operatives.
- Built and deployed automated data quality assurance and data reporting pipelines.
- Curated interactive visualizations and analytical reports on the status of the company's data for investors and clients.
Machine Learning Scientist2016 - 2016Gyana
Technologies: PostGIS, PostgreSQL, APIs, Flask, Scikit-learn, Keras, TensorFlow
- Built models using gradient boosting/random forests/deep learning with a heavy emphasis on time-series geospatial data, combining multiple diverse data streams into features.
Data Scientist | Back-end Developer2016 - 2016Equidam
Technologies: Flask, Gunicorn, NGINX, Bash, Linux, Ansible, PostgreSQL, Python, Cassandra, Docker
- Built fully automated crawling systems using Docker, Cassandra, Python, and PostgreSQL.
- Deployed with Ansible.
- Automated raw data processing.
- Developed various ML frameworks for modeling and predictive analytics on top of this raw data.
Graduate Researcher2015 - 2016Santa Fe Institute
Technologies: SQLite, NetworkX, Pandas, Python
- Performed complex modeling and analysis of a terabyte-sized dataset of patents to gain insights into the process of technological innovation.
Graduate Researcher2015 - 2015Institute for New Economic Thinking — University of Oxford
Technologies: SQLite, SQL, NetworkX, Pandas, Python
- Analyzed a corpus of patent data through advance means to publish in EPJ Data Science journal: “Understanding Technological Change from the Map of Capabilities.”.