Experience

• 

  Niit limited

  Jun 2018 - Jul 2018

  Software engineer intern

  • Architectured an existing Assessment System tool with React for a new feature to improve customer adoption. • Built a web scrapper using UiPath RPA tool.

• 

  Niit limited

  Jun 2019 - Jul 2019

  Software engineer intern

  • Build a Facebook and Twitter sentiment analysis tool. Sentiments of reviews, posts and comments can be calculated and visualized to help the team plan and produce effective campaigns that can captivate their target audience.• Data extraction was made possible by Facebook’s graph API and Tweepy.Tech Stack: Bidirectional Encoder Representations from Transformers (BERT), Django, MongoDB

• 

  Unitedhealth group

  Aug 2020 - Jun 2022

  Data scientist

  𝟏. 𝐓𝐫𝐚𝐧𝐬𝐟𝐨𝐫𝐦𝐢𝐧𝐠 𝐈𝐧𝐜𝐢𝐝𝐞𝐧𝐭 𝐌𝐚𝐧𝐚𝐠𝐞𝐦𝐞𝐧𝐭 𝐢𝐧 𝐈𝐓 𝐒𝐮𝐩𝐩𝐨𝐫𝐭Developed and deployed an advanced automated system using Machine Learning to streamline IT Support Ticket classification. Key features include:• Real-time and Historical Categorization: Implemented automated categorization for both current and past incidents, ensuring fast and accurate classification.• Resolution Prediction: Created predictive models to forecast resolutions, enabling proactive management and quicker problem-solving.• Resolution Tagging: Automated the tagging process for resolutions to streamline tracking and retrieval.𝐁𝐞𝐧𝐞𝐟𝐢𝐭𝐬 𝐀𝐜𝐡𝐢𝐞𝐯𝐞𝐝:• Faster Categorization: Enabled proactive measures with swift, accurate incident classification.• Increased Efficiency: Automation improved precision, saving approximately 4-5 Full-Time Employees (around $240,000 annually).𝐃𝐞𝐩𝐥𝐨𝐲𝐦𝐞𝐧𝐭: Utilized HCC for orchestration, Docker for containerization, and CI/CD with Jenkins for continuous integration.𝟐. 𝐄𝐧𝐡𝐚𝐧𝐜𝐢𝐧𝐠 𝐈𝐓 𝐏𝐨𝐫𝐭𝐚𝐥 𝐰𝐢𝐭𝐡 𝐀𝐝𝐯𝐚𝐧𝐜𝐞𝐝 𝐒𝐞𝐚𝐫𝐜𝐡 𝐂𝐚𝐩𝐚𝐛𝐢𝐥𝐢𝐭𝐢𝐞𝐬Developed a self-service IT portal with improved search functionality, integrating semantic search and Elasticsearch re-ranking. 𝐁𝐞𝐧𝐞𝐟𝐢𝐭𝐬 𝐀𝐜𝐡𝐢𝐞𝐯𝐞𝐝:• Advanced Search Capabilities: Enhanced search functions resulted in a reduction of Average Handling Time (AHT) by 20 seconds, boosting operational efficiency.𝟑. 𝐀𝐝𝐯𝐚𝐧𝐜𝐢𝐧𝐠 𝐇𝐞𝐚𝐥𝐭𝐡𝐜𝐚𝐫𝐞 𝐃𝐚𝐭𝐚 𝐌𝐚𝐧𝐚𝐠𝐞𝐦𝐞𝐧𝐭Enhanced X-ray Image Processing:Applied OpenCV and scikit-image to pre-process X-ray images, ensuring clarity and standardization for accurate text extraction.Automated PII Detection and Redaction:Developed algorithms for detecting and redacting Personally Identifiable Information (PII) from X-ray reports, ensuring HIPAA compliance and supporting public health initiatives. Show less

• 

  Uniting

  Sept 2022 - Dec 2023

  Data scientist (via uts-uniting collaborative project)

  𝐄𝐧𝐡𝐚𝐧𝐜𝐢𝐧𝐠 𝐄𝐥𝐝𝐞𝐫𝐥𝐲 𝐇𝐨𝐦𝐞 𝐂𝐚𝐫𝐞 𝐰𝐢𝐭𝐡 𝐏𝐫𝐞𝐝𝐢𝐜𝐭𝐢𝐯𝐞 𝐑𝐢𝐬𝐤 𝐀𝐬𝐬𝐞𝐬𝐬𝐦𝐞𝐧𝐭 𝐚𝐧𝐝 𝐃𝐚𝐬𝐡𝐛𝐨𝐚𝐫𝐝 𝐌𝐨𝐧𝐢𝐭𝐨𝐫𝐢𝐧𝐠• Predictive Model Development: Created a model to identify and understand the complex factors leading to health decline among elderly clients. This model leverages data-driven insights to anticipate and address potential health issues before they escalate.• Timely Interventions: The model is designed to facilitate early interventions, personalized care plans, and comprehensive support systems. This approach aims to extend the duration of safe and comfortable living at home.• Monitoring Dashboard: Developed a dedicated dashboard using Power BI to track the model's performance and provide actionable insights. This tool is essential for healthcare practitioners, caregivers, and administrators, offering a streamlined way to monitor and manage the well-being of elderly clients in HCPs.𝐈𝐦𝐩𝐚𝐜𝐭 𝐚𝐧𝐝 𝐆𝐨𝐚𝐥𝐬:Hospitalization Risk Reduction: Aims for a 30% reduction in hospital readmissions by enhancing proactive care and support.Quality of Care Improvement: Represents a significant advancement in elderly care, ensuring that individuals can age in place with dignity, security, and optimal support. Show less

• 

  University of technology sydney

  Jan 2023 - Jun 2024

  𝐏𝐫𝐞𝐝𝐢𝐜𝐭𝐢𝐧𝐠 𝐀𝐮𝐭𝐨𝐦𝐚𝐭𝐢𝐨𝐧 𝐑𝐢𝐬𝐤𝐬: 𝐔𝐩𝐬𝐤𝐢𝐥𝐥𝐢𝐧𝐠 𝐚𝐧𝐝 𝐑𝐞𝐬𝐤𝐢𝐥𝐥𝐢𝐧𝐠 𝐒𝐭𝐫𝐚𝐭𝐞𝐠𝐢𝐞𝐬 𝐔𝐬𝐢𝐧𝐠 𝐋𝐋𝐌𝐬• Implemented a simulation model using Gemini, LangChain, and RAG, which identified 25% of occupations as highly susceptible to automation, guiding strategic workforce planning.• Developed an LLM-based application that creates detailed knowledge graphs from over 500 documents using Gemini Vision and Pro models to convert them into structured JSON data.• Enhanced document processing accuracy using LangChain’s MapReduce and Refine Chains, to iteratively refine outputs.• Deployed the solution using LangServe, seamlessly integrating JSON data into Neo4j for visualization and analysis.• Leading an in-depth analysis to tailor up-skilling/re-skilling recommendations in response to the findings.𝐄𝐥𝐝𝐞𝐫𝐥𝐲 𝐇𝐨𝐦𝐞 𝐂𝐚𝐫𝐞 𝐑𝐢𝐬𝐤 𝐀𝐬𝐬𝐞𝐬𝐬𝐦𝐞𝐧𝐭 • Predicted hospitalization risks for elderly Australians to help them stay at home longer, targeting a 30% reduction in readmissions by identifying high-risk clients and enabling proactive care interventions.• Enhanced overall model accuracy by 10% through the fine-tuning of a spaCy NER model for precise disease identification from patient records and by utilizing BertTopic for in-depth thematic analysis.• Enhanced the recall of risk assessments by 15% using Autoencoder and ANN to classify and rank risk levels effectively.• Designed and implemented a PowerBI dashboard for continuous monitoring of elderly clients in home care packages. Show less

  • Research Assistant

    Sept 2022 - Jun 2024

  • Academic Board Representative (TD School)

    Jan 2023 - Nov 2023
• 

  Foboh

  Oct 2024 - now

  Ai engineer

  Developing FOBOH Intelligence, an AI-driven co-pilot for optimizing order management in the Food & Beverage wholesale industry.