Keanu Ghorbanian Email & Phone Number

United Kingdom

London

- Building AI Agents to provide guidance and assistance to colleagues and customers, integrated into existing workflows.- Building forecasting models for financial forecasting of loans and applications

London, United Kingdom

Keywords: Python, C++, Data Analysis, Statistics, Machine Learning, High-Performance Computing (HPC)• Research - Measurement of physics processes in the LHC is a data science. Data cleaning, filtering, fitting are daily tasks, as well as working with large, complex, and messy datasets, and implementing statistical and machine learning methods.• Framework Development - Engineered a framework for high-volume particle data analysis for TB+ datasets and Monte Carlo simulations within the… Show more Keywords: Python, C++, Data Analysis, Statistics, Machine Learning, High-Performance Computing (HPC)• Research - Measurement of physics processes in the LHC is a data science. Data cleaning, filtering, fitting are daily tasks, as well as working with large, complex, and messy datasets, and implementing statistical and machine learning methods.• Framework Development - Engineered a framework for high-volume particle data analysis for TB+ datasets and Monte Carlo simulations within the in-house CERN C++/Python analysis libraries. Vastly improvemed speed and accuracy by leveraging distributed compute and integrating visualisation.• Detector upgrade - Led upgrade efforts for the monitoring software of the inner detector trigger in the ATLAS experiment. Successfully recreated legacy algorithms in Python and C++, ensuring real-time monitoring of hardware data-taking. This initiative significantly enhanced the research teams’ abilities to detect and debug issues promptly. Show less

London, United Kingdom

I tutor A-Level Physics, Chemistry and Mathematics privately to students aged 16 - 18

London

Keywords: Finance, Python, machine learning, feature selection, Azure• Internship at a £4 billion+ commodities investment fund.• Backengineering competitor investments - Using timeseries stock data, I developed a feature selection framework to predict investments based on a portfolio’s P&L. This involved development of custom loss and selection algorithms to enhance accuracy of investment predictions.• Allowed company to not only predict competitor’s investments, but quantitatively… Show more Keywords: Finance, Python, machine learning, feature selection, Azure• Internship at a £4 billion+ commodities investment fund.• Backengineering competitor investments - Using timeseries stock data, I developed a feature selection framework to predict investments based on a portfolio’s P&L. This involved development of custom loss and selection algorithms to enhance accuracy of investment predictions.• Allowed company to not only predict competitor’s investments, but quantitatively define the predictability and diversity of their own internal portfolios; empowering the company to make better decisions about fund allocation and portfolio management. Show less

Keywords: Python, SQL, APIs, Azure, NLTK, tensorflow• Internship at one of the UK’s oldest and largest charities.• Weather data integration - Utilized SQL server, and Python to integrate and analyse RNLI’s incident weather data on Microsoft Azure with online weather APIs. This revealed under-reporting of weather conditions by lifeguards, and enabled supported management in validation and funding allocation for lifeboat stations.• Incident summary recategorisation - Used NLTK and… Show more Keywords: Python, SQL, APIs, Azure, NLTK, tensorflow• Internship at one of the UK’s oldest and largest charities.• Weather data integration - Utilized SQL server, and Python to integrate and analyse RNLI’s incident weather data on Microsoft Azure with online weather APIs. This revealed under-reporting of weather conditions by lifeguards, and enabled supported management in validation and funding allocation for lifeboat stations.• Incident summary recategorisation - Used NLTK and Tensorflow to re-categorise incident lifeguard reports, vastly improving internal statistical tasks. Show less

Dongguan, Guangdong, China

I taught A-Level Physics in the international department of Humen Foreign Language School, Dongguan.I also taught English, Mathematics and Mixed Sciences at middle-school level.

Sao Carlos, Brazil

I looked to confirm a proposition that the combined effect of both old andyoung Supernova Remnants (SNRs) can efficiently accelerate cosmic rays up to a transition in the spectrum named 'the knee'. Under this scenario, the spectral hardening of cosmic ray hadrons can be interpreted as the transition between the two populations of sources. I tested this scenario in view of the latest SNR catalogue provided by the Fermi-LAT experiment.