Diya Nair

Experience

Eon Media

Jan 2025 - April 2025

Automated video encoding and metadata scraping using YT-DLP , FFmpeg, Python, Selenium, BeautifulSoup, and Postman, streamlining media processing workflows, improving code reliability by 25%.
Tested Eon products to ensure enhanced user experience, optimized automation testing for Eon AI products, and annotated datasets to improve logo detection.
Conducted AI research to improve logo and scoring detection using OpenCV and Librosa
Assisted AI/ML engineers to build an end-to-end pipeline for audio content analysis, integrating logo/brand detection using YOLOv8, voice recognition, and category/summary generation using fine-tuned LLMs and NER.
Facilitate Sprint Planning and backlog refinement, optimizing JIRA to reduce backlog size by ~30% and streamline project delivery.

William Osler Health System

July 2023 - August 2023

Assisted with daily diagnostic imaging routines and processed over 40 x-rays and 15 patients per day
Enhanced patient experience by receiving over 99% positive feedback from patients and the radiology team.

C++, Python, Wireguard, Proxmox Server, pfSense Firewall, Dell PowerEdge R730 Server

Setting up a home lab Library Manager using C++ and Python to experiment with data structures like linked lists, trees, heaps, and hash tables. Implemented backend logic for book cataloging, user operations, and other features. Storing and managing data using PostgreSQL.
Exploring microservices architecture for authentication and user management via REST APIs.

Python, GUI, LangChain, OpenAI (ChatGPT API), BeautifulSoup, Streamlit, Chroma

Developed an LLM-based chatbot capable of scanning websites and providing interactive, context-aware answers
Used OpenAI's ChatGPT models for generating conversational responses and hosted on Streamlit.
Integrated Beautiful Soup for data extraction and Chroma Vector Store for data storage and retrieval.

Python, Pygame

Developed a custom version of the popular Dino game using Python and Pygame, with dynamic GUI gameplay
Utilised object-oriented programming to design custom classes to control sprite behaviour
Implemented scoring logic, obstacle spawning, and character control.

RobotC, Lego EV3, PID

Designed an autonomous ambulance robot with Lego EV3 to navigate a maze by simulating road conditions
Integrated motors, ultrasonic, color, touch, sound, and gyro sensors for obstacle detection, and navigation.
Applied a PID controller system to enable accurate 90° turns and prevent overshooting.