Hey, I'm Kumkum

Student @ UOFT| AI Enthusiast | Software Developer

Passionate about leveraging AI and software development to solve complex challenges. Experienced in building predictive models, interactive web applications, and data-driven solutions.

Hero Section

Experience

Student Researcher, University of Toronto

Student Researcher, University of Toronto

I contribute to a research initiative that explores how emotions impact decision-making, with a particular focus on long-term memory. My work centers on analyzing language data to understand the relationship between emotional word norms and their usage in text. I develop scripts that examine how emotional characteristics of words relate to their log frequency and position within a corpus. As part of this project, I design and implement regression models that correlate emotional norms with linguistic features. This analysis helps uncover how emotionally charged language influences memory and decision-making processes, offering insights into the cognitive mechanisms behind emotional interpretation in language.

Software Developer, Medium AI

Software Developer, Medium AI

At Medium AI, I was actively involved in the full-stack development of a medical scribe platform, leveraging React and JavaScript to optimize user experience and scalability. In collaboration with the AI research team, I contributed to the design and implementation of a DirectRAG model that transformed transcripts into precise, medical-grade documentation. This role allowed me to merge advanced AI technology with intuitive design, addressing real-world challenges in medical documentation

Data Engineer, Northbridge Insurance

Data Engineer, Northbridge Insurance

At Northbridge Insurance, I focused on developing AI-driven solutions, particularly in the realm of chatbot and data automation. I built an intelligent chatbot designed to translate natural language into SQL queries, using open-source large language models (LLMs) and advanced prompt engineering techniques. By integrating Retrieval-Augmented Generation (RAG), the chatbot could clarify user inputs and offer adaptive responses, enabling seamless interaction with complex databases. Leveraging the Cortex API, I enhanced the chatbot's precision by 40%, ensuring it provided contextually accurate, data-driven answers. This innovation improved internal data accessibility, empowering users to make informed decisions efficiently. Additionally, I deployed these advancements through intuitive web applications with Streamlit and React, optimizing user experience and workflow productivity

Internal Outreach Director, Illuminate

Internal Outreach Director, Illuminate

As an Internal Outreach Director, I lead initiatives to engage and inspire students within our organization, fostering a collaborative and inclusive environment. I coordinate and promote events that bridge interdisciplinary interests in computer science, AI, and ethics, helping students connect with opportunities, gain insights, and develop a network for career and academic growth. My role involves liaising with team members to organize impactful events, streamline communications, and build lasting connections within the community.

About Me

About Me

I am a Computer Science, Cognitive Science and Mathematics student at UofT. I have a strong foundation in programming, data science, and machine learning, combined with a passion for AI applications. With hands-on experience in data engineering, web development, and cognitive modeling, I excel at creating solutions that bridge technology and practical needs. I am currently working as a Student Research in the field of AI and Cognitive Science.

Skills

Front-End Development

Proficient in React and Streamlit, building interactive, responsive UIs. Improved user experience by 25% through intuitive design and functionality integration.

Back-End & API Integration

Skilled in REST APIs, SQL, and Azure Data Factory (ADF), streamlining data pipelines and enhancing database performance with Snowflake.

Data Science & Machine Learning

Experienced in PyTorch, TensorFlow, scikit-learn, and advanced data visualization with Power BI, specializing in predictive modeling and data analysis.

Large Language Models (LLMs) & Chatbots

Expertise in open-source LLMs, prompt engineering, and Retrieval-Augmented Generation (RAG), achieving a 40% improvement in chatbot answer precision.

Data Visualization

Enhanced report clarity and engagement using Power BI and visual analytics, delivering actionable insights through custom dashboards for decision-making.

System Design & Optimization

Experience with system design for projects like a solar-powered heating system, optimizing water heating and purification, and designing scalable back-end systems.

Programming Languages

Proficient in Python, Java, C++, SQL, and MATLAB, with experience in leveraging these languages for machine learning, data analysis, and software development projects.

Research

Experienced with computational modeling, analyzing cognitive processes, optimizing model parameters, and validating findings through simulations, with experience in interdisciplinary research and creating clear visualizations.

Concpet Learning Model

Python • PyClarion • Cognitive Modeling • SNN

Developed a cognitive model for concept learning using the Clarion architecture, explaining human concept processing and categorization by integrating sensory, and linguistic representations through an associative merging mechanism.

View in GitHub

DRM Sleep Simulation

Python • LSTM • Neural Networks • Cognitive Science • Sleep Research

Designed and implemented a Python-based simulation model utilizing LSTM neural networks to analyze the effects of sleep on DRM false memory formation, successfully replicating empirical study findings.

View in GitHub

F1 Predictive Model

Python • RNN • LSTM • Machine Learning • Data Analysis

Developed a machine learning model using RNN and LSTM to predict F1 driver lap times and positions, achieving 80% accuracy and predictions within 3 seconds of target lap times.

View in GitHub

Item Response Theory (IRT) Model

Python • Machine Learning • Statistical Analysis

Enhanced an IRT model for individualized learning assessments, achieving a 90% improvement in accuracy by optimizing model parameters and algorithms.

View in GitHub

Image Filter Server

C++ • Web Development • Image Processing • Server Architecture

Built a C++ server application enabling users to upload photos and apply filters via a user-friendly web interface, optimizing the image processing workflow for enhanced performance.

View in GitHub

Recipe Search Software

Java • REST APIs • Collaborative Development • User Interface Design

Developed a recipe search application in Java with a collaborative team, integrating APIs and user-driven contributions to boost daily active users by 40%.

View in GitHub