Software Development & Machine Learning Engineer — LLM Integration • Frontend Dev • CI/CD Automation

📍 San Francisco, California · 💼 vibha.bhavikatti@gmail.com · 🌐 GitHub · LinkedIn

🎓 CS Graduate Student @ San José State University
🧠 Passionate about ML, LLMs, Backend Systems, and DevOps
🚀 Seeking SWE/MLE Internships and New Grad roles
🏆 1st place 🇦🇪 + 6th place 🌍 in the NASA-JAXA KIBO Robot Programming Competition

Technical Advisor Intern - GenAI

Scale AI (Aug 2025 - Present)

• Evaluated LLM reasoning and failure modes at scale; contributed structured insights to improve generative AI reliability and integration into developer workflows.
• Designed prompt variations and agent-style evaluation tasks to analyze model behavior under edge cases and ambiguous user input.
• Developed simulation-based pipelines and custom metrics to automate evaluation of LLM outputs, enabling large-scale analysis of AI performance; used Selenium and Locust for task automation and workflow testing.

Information Technology Intern
Emirates Fast Food Co. (McDonald’s UAE) (Jun 2023 – Jul 2023)

• Resolved IT support tickets via the Ivanti service desk and automated database workflows, improving system uptime by∼10%.
• Assisted with database management, automation tools, and product pricing workflows.

Teaching Assistant
American University of Sharjah – Multiple Semesters

• Conducted Programming I lab sessions for 30+ students
• Graded Programming I and Digital Systems courses under 3 professors

GDSC Non-Technical Core Lead
Google Developer Student Club (AUS) (Jun 2023 – Jul 2023)

• Mentored 100+ students in Programming I, Programming Languages and Digital Systems; provided hands-on support in lab sessions and code reviews. Assisted professors with grading, documentation, debugging, and course support across 4 semesters.

YOLOv3/YOLOv8 + GCP-based web app for object detection with geo-mapping and metadata.

🥇 mAP@0.5 = 96.1% | Image de-duplication | Real-time inference

Flask NLP app using RoBERTa, LLaMA 3, PoliticalBiasBERT, and LLM summarization via Hugging Face.

🌐 Web scraping | Real-time REST APIs | LLM-based summaries

Time-series forecasting using LSTM/Transformer models on sentiment-enriched news.

🧠 Engineered sentiment/polarity features | 0.60 F1-score (beating benchmark scores with lesser computational resources)

Full-stack app using React + Firebase to deliver daily coding problems with progress tracking.

🧪 CI/CD via GitHub Actions | Dockerized | Cypress + Vitest tests

CNN model detecting COVID-19 from X-ray images.

🥈 4th Place | 97.21% accuracy | Built in 48 hours

Languages:
Python, C++, JavaScript, Java, SQL, R, Rust, TypeScript, MATLAB, HTML/CSS, Bash

Frameworks:
PyTorch, TensorFlow, Flask, React, scikit-learn, Keras, MapReduce, Hadoop, Qt

Cloud & DevOps:
Google Cloud (GCP), AWS (EC2, Lambda, S3), Docker, Kubernetes, GitHub Actions, CI/CD

Testing & Automation:
Vitest, Cypress, JUnit, Selenium, Postman, UAT, Load Testing

Big Data:
Hadoop, Spark, Hive, ETL pipelines, Data Modeling, Ingestion, Distributed Systems

• 📧 Email: vibha.bhavikatti@sjsu.edu
• 🔗 LinkedIn: linkedin.com/in/vibha-bhavikatti
• 💻 GitHub: github.com/vibhab4

✨ Thanks for stopping by!