Rayhaan Pirani

Experience

• 

  GGK Tech

  Dec 2017 - Mar 2018

  Trainee Software Engineering Intern

  - Worked on the development of apps using C#/.NET.- Undergone training in Node.js and Angular.- Attended 12 training sessions.

• 

  DBS Bank

  Jul 2018 - Aug 2021

  - Developed an integrated development workbench by creating an automated data engineering pipeline in Hadoop, Spark, Hive, and Python with a Qlik dashboard and a responsive Web 2.0 interface to help institutional banking relationship managers assess incomes and balances and compare them with projected incomes and balances to make data-driven financial decisions.- Designed an early warning system that detected over 95% of non-performing SME loans ahead of 3 months of credit stress, safeguarding more than $700 million (SGD).- Implemented agile methodologies and ML-Ops, partnering with diverse teams, and utilizing Git, Subversion, JIRA, and Confluence for streamlined project management. Show less - Fine-tuned an NLP-driven chatbot using SVMs, spaCy, and NLTK, achieving 94% accuracy for Cantonese-speaking Hong Kong corporate clients, addressing 80% of inquiries and substantially reducing support calls.- Designed and maintained ETL data workflows in Hadoop, Hive, and Spark, employing Python, R, Java, SQL, and Jupyter on Cloudera.- Implemented agile methodologies and ML-Ops, partnering with diverse teams, and utilizing Git, Subversion, JIRA, and Confluence for streamlined project management.- Architected 8 Qlik dashboards aiding SME banking in Asia, advancing digitization, customer network analysis, loan oversight, and balance sheet strategies.- Built an automated web scraping pipeline in Python using BeautifulSoup and Selenium to obtain data from a Hong Kong government data repository for companies.- Awarded the Rising Star award in August 2019 and Best Employee award in June 2021 based on my performance. Show less

  • Application Developer - Data Science (Analyst)

    Sept 2020 - Aug 2021

  • Application Developer - Data Analytics (Senior Officer)

    Sept 2018 - Sept 2020

  • Technical Associate - Training Program

    Jul 2018 - Sept 2018
• 

  University of Windsor

  Oct 2021 - Aug 2023

  Graduate Teaching Assistant

  - Collaborated with the course professor to ensure student success and overcome coursework challenges.- Organized and planned lab sessions for around 30 students for algorithms and programming in C.- Graded assignments and examinations for 20-25 students and provided relevant feedback.- Conducted weekly office hours to answer questions and guiding students.

• 

  Immigration, Refugees and Citizenship Canada / Immigration, Réfugiés et Citoyenneté Canada

  Jan 2022 - Mar 2024

  NLP Research Assistant

  - Developed an AI recommendation system using a fine-tuned sBERT language model on Canadian immigration data, aimed at aiding settlement agencies. This system provides personalized recommendations to caseworkers based on client questionnaires, facilitating the support for newcomers to Canada.- Secured a significant funding of $100 million (CAD) over the next 3 years from the Immigration, Refugees and Citizenship Canada (IRCC) for a pivotal project among 78 initiatives, exploring the potential of AI in crafting effective settlement strategies for newcomers. Show less

• 

  Liberty

  May 2022 - Dec 2022

  Performance Analyst & Engineering Co-op

  - Implemented analytics solutions with the performance and engineering teams, utilizing Python, advanced Excel, pivot tables, and Tableau to optimize wind turbine performance.- Led aerodynamic studies on blade angle adjustments across 89 turbine sites, harnessing fluid dynamics simulations, and realizing operational cost savings of over $575,000 (USD) per site.- Utilized anomaly detection techniques, like One-Class SVM and Isolation Forest, on diverse weather datasets extracted from various APIs, creating predictive models to pinpoint ice accumulation and abnormal turbine operations. Achieved a notable reduction in downtime, averting around 150 hours per turbine.- Automated data mining for wind speed insights from multiple sources, employing APIs and web scraping with BeautifulSoup and Selenium, enhancing the understanding of wind patterns for informed wind turbine site selection. Show less

• 

  Vector Institute

  Sept 2023 - Dec 2023

  Machine Learning Associate

  - Employed NLP techniques and fine-tune language models like BioBERT and ClinicalBERT to extract key insights from patient texts, such as chief complaints and standardized medical codes (SNOMED-CT and ICD-10).- Utilized LLMs, including GPT-4 and Llama 2, to generate human-like synthetic training data.- Developed an in-house LLM infrastructure on AWS, specifically fine-tuned on healthcare data, to tackle advanced medical reasoning tasks like disease detection, classification, and syndromic surveillance processes. Show less