Smart Product Search
Multimodal semantic search across 12K items using vision–text embeddings.
- 0.65 MRR, 78% Hit@10 with dual-encoder + cosine similarity on vector storage.
- REST API with Flask; evaluation suite for relevance & latency.
Software Engineer · MS CS @ Georgia Tech (4.0/4.0)
Actively seeking full-time Software Engineering opportunities starting in 2026. Get in touch with me here!
I am a second-year Master’s student in Computer Science at Georgia Tech. Most recently, I interned at Amazon on the Books Recommendations team, where I developed an LLM evaluation framework for qualitative analysis of book recommendations.
Previously, I worked as a Software Engineer at JPMorgan Chase, contributing across the full stack—frontend, backend, CI/CD pipelines, cloud infrastructure, and databases while owning and delivering scalable features for Trade Finance applications.
I earned my undergraduate degree in Computer Engineering from the University of Mumbai, graduating with a GPA of 9.80/10. I’ve also completed internships at X80 Security (building security analytics dashboards and AWS infrastructure) and JioSaavn (optimizing performance for the ads platform using SQL, Node.js, and React).
I iterate fast and love to take ideas from concept to production. Interests include LLMs, backend development, MLOps, cloud computing, and full-stack engineering.
Languages: C++, Java, TypeScript, Python, Go, JavaScript, HTML, CSS
Frameworks & DBs: React, Spring Boot, Django, Node.js, Express.js, Redux, GraphQL, MySQL, PostgreSQL, MongoDB, Redis
Cloud & DevOps: AWS, Google Cloud (GCP), Docker, Kubernetes, Jenkins, CI/CD, EKS, EC2, Lambda, S3
AI & Data: LLMs, RAG, TensorFlow, PyTorch, NumPy, Kafka, Spark, Hadoop
Practices: Git, Linux, Agile, SDLC, UI/UX
AWS Developer Associate — Validation #286EJ8M2TB1QQC5K
Multimodal semantic search across 12K items using vision–text embeddings.
Interactive tool for teaching protocols, packet flow, and synchronization.
All-in-one platform for tutors & students to manage batches, lectures, tests, and materials.
Raspberry Pi system infers infant emotions from synchronized audio + video streams.
Client-side Intel 8086 emulator with interactive interpreter and highlighting.
tghone3@gatech.edu · +1 (404) 642-3760 · Atlanta, GA