Tirth Gada

Software Developer Intern at Trilogy Innovations (formerly Codenation), enhancing a multi-agent platform with advanced file handling, caching, and AI-driven solutions.

• Enhanced multimodal file handling for an MCP-multi agent platform on Claude 3.5 Sonnet, integrating presigned URLs for efficient data access.
• Optimized system performance by implementing LRU SSD caching, reducing latency by 60% and enabling efficient file referencing for inbuilt agents.
• Developed image captioning and indexing capabilities for LLM context, leveraging OpenAI and AWS OpenSearch DB with chunking for enhanced search.
• Scaled Server-Sent Events (SSE) using Redis on ECS, integrating GitHub and G-drive via OAuth2 for seamless data flow and resource management.

Technical Intern at Siemens, developing industrial safety solutions and internal platforms, applying machine learning and web development skills.

• Developed an industrial safety solution for CCTV, implementing real-time object detection using YOLO for enhanced security monitoring.
• Implemented WebSocket-based server-side scripting with Node.js, enabling real-time communication of safety violation alerts.
• Designed and deployed a Restricted Zone Detection System, providing real-time alerts to enforce safety protocol adherence.
• Built the frontend for an internal chatbot platform using React and the organization's IX framework for styling and user experience.

Machine Learning Intern at Qrata, an AI & Technology driven talent acquisition firm, recognized for excellent performance in developing talent matching systems and optimizing data processing.

• Led the development of 4+ Django microservices, including semantic extraction, to improve talent ranking for over 20,000 profiles.
• Optimized data ingestion efficiency by 80% through multi-threading for concurrent file fetching and automated processes with a cron scheduler.
• Leveraged NLP techniques, LangChain, and OpenAI GPT to generate high-quality resume embeddings for enhanced search capabilities.
• Developed an intelligent talent matching system using K-Means clustering and a smart search bar integrated with VectorDB.

Mentored 8 students in object detection and deep learning concepts, providing guidance and support.

• Guided 8 students through object detection (YOLO), image classification, and GAN paper implementation.
• Developed educational resources and assignments on Neural Networks and Deep Learning, actively supporting student learning.

Coordinated web initiatives for the Student Alumni Relations Cell at IIT Bombay, managing web portals and engaging over 60,000 alumni and 12,000 students.

• Revitalized the frontend of 5+ web portals using VantaJS and GSAP, enhancing user experience and engagement.
• Engineered the backend for a crypt hunt game platform using Django, successfully engaging over 2,000 participants.
• Managed ground operations for Alumination, a flagship fest, serving over 2,000 students.
• Orchestrated a bicycle donation campaign for underprivileged communities, ensuring efficient logistical execution.

Supervised students in developing a news sentiment classification model.

• Oversaw 10+ students in the development of a news sentiment classification model, utilizing Natural Language Processing techniques.

Developed a Generative Adversarial Network (GAN) to create high-quality cartoon faces and colorize images, implementing a research paper.

Developed a Neural Machine Translation (NMT) system to translate English to Hindi, achieving high accuracy.

Participated in the International Quant Championship, securing a top rank and receiving a research consultant offer, focusing on developing quantitative trading strategies.

Participated in a hackathon, developing a fraud detection system using various ML models and extensive data analysis techniques.

Analyzed Spotify dataset to identify optimal music clusters using K-Means and PCA, leveraging 8+ audio features.

Developed an AI-powered Sudoku solver that extracts boards from images and solves them using a backtracking algorithm.

Streamlined registration and authentication processes for IIT Bombay's Student Alumni Relations Cell across multiple web portals.

Developed advanced data structure and algorithm solutions for efficient data processing and recommendations, employing OOPS principles.

Pioneered a web marketplace for IIT Bombay students, inspired by OLX, facilitating seamless buy/sell of used items and incorporating real-time chat.

Achieved World Rank 154 out of 34,000+ participants and offered a position as a Research Consultant.

Awarded a Letter of Recommendation from the Director for excellent performance during Machine Learning Internship.

Ranked amongst the top 2 percentile in the JEE Advanced examination among 0.18+ million aspirants.

Achieved 99+%ile in the JEE Main examination in both attempts among 0.95+ million candidates.

Obtained an impressive score of 99.938%ile in the Maharashtra Common Entrance Test (MHT-CET).

Cleared KVPY-SX examination conducted by IISc & secured a Global Rank 55 in SOF IMO Olympiad.