JD.

02 : Experience / Learning through the work

Professional
experience.

Teaching, engineering, and everything learned along the way. The roles where I've put curiosity into practice.

— Present

Part-time
Baltimore, Maryland · On-site

The Johns Hopkins University

Teaching Assistant

EN.553.171: Discrete Mathematics, Department of Applied Mathematics and Statistics.

  • Create section worksheets that strengthen students’ understanding of course concepts.
  • Grade homework and provide feedback on mathematical reasoning and correctness.
  • Hold weekly office hours to help students work through logic, combinatorics, modular arithmetic, graph theory, and other course topics.

Tools LaTeX · Overleaf

—

Internship
United States · Remote

Optivoy

AI Engineering Intern

  • Built and deployed custom Python MCP servers for Gmail, Microsoft Graph, LiteAPI, and weather services, enabling autonomous retrieval of travel history, loyalty memberships, live flight options, and destination forecasts at arrival.
  • Developed an interactive LLM-powered flight optimization assistant with access to flight search, past trips, calendar information, and loyalty data. Travelers could refine recommendations conversationally around price, schedule, layovers, rewards, and personal preferences.

Tools Python · FastMCP · Anthropic Claude

—

Internship
Bengaluru, India

CollegeSource

Intern

  • Automated the creation and maintenance of a structured database of 436+ U.S. universities, scraping and aggregating data with Python and BeautifulSoup.
  • Collected and organized acceptance rates, student–faculty ratios, average class sizes, tuition and fees, graduation rates, SAT/ACT ranges, popular majors, rankings, and notable alumni.
  • Exported cleaned datasets into dynamically formatted Excel workbooks using OpenPyXL, saving 30+ hours of manual research and data entry.

Tools Python · BeautifulSoup · OpenPyXL

—

Internship
Bangalore Urban, India · On-site

MatryxSoft Tech

Software Engineering Intern

  • Trained a real-time image recognition model using YOLOv8, PyTorch, and Roboflow to classify 2,000+ beverage categories.
  • Engineered a Python and YouTube Data API pipeline to extract video captions and structure content for website generation, reducing manual data collection by 10+ hours.
  • Trained and evaluated food detection and classification models using TensorFlow (Keras), Ultralytics YOLO, and ResNet, achieving 87% accuracy across 100+ dishes and 300+ ingredients.
  • Built an OpenCV solution to detect vacant shelf regions and estimate maximum bottle capacity using bounding box dimensions and spatial constraints.

Tools Python · YOLOv8 · PyTorch · Roboflow · TensorFlow · OpenCV