Greha Shah

- Building cloud-native services on GCP to support scalable merchant integration and commerce capabilities. - Contributing to the design and deployment of distributed backend systems handling high-throughput workflows. - Driving improvements in reliability, observability, and performance across cloud infrastructure. - Partnering with senior engineers to ship secure, scalable solutions used by global merchants.

Grader for the course: CSE 360 (Intro to Software Engineering)

• Built an LLM-powered merchant insights chatbot delivering natural language answers up to 17× faster than Kibana, cutting query time from 6 minutes to 22 seconds for 1M+ records and improving decision-making efficiency. • Built a 2-stage LLM pipeline converting questions to SQL for BigQuery, then using another LLM to deliver clear, contextual responses. • Benchmarked 7 LLMs (15 parameters) and added error handling for 5xx/non-2xx responses, achieving 99.9% production uptime.

Developed and optimized machine learning models for treatment recommendations. Collaborated with psychologists and engineers to preprocess data, train models, and evaluate performance. Designed and implemented an intelligent chatbot to assist with patient queries and support. Integrated natural language processing (NLP) techniques to improve chatbot understanding and response accuracy. Conducted user testing and iterated on chatbot features based on feedback.

Spearheaded a community of passionate students, driving innovation with Google's technologies. Organized workshops, hackathons, and speaker sessions to foster collaborative learning. Coordinated with industry professionals to provide mentorship and real-world insights. Grew the club's membership and impact, facilitating numerous technical projects and initiatives.

At DJ UNICODE, I embarked on a transformative journey in web development. In my first year as a mentee, I developed a strong foundation in front-end and back-end technologies, mastering tools and languages essential for web development. Over the next two years, I transitioned to a mentor role, where I led projects, conducted workshops, and provided guidance to new members. This role allowed me to enhance my leadership and technical skills, fostering a collaborative and innovative learning environment within the team.

Created a responsive website using React.js and Next.js, boosting mobile traffic by 45% and reducing bounce rates by 30%. Enhanced data retrieval and API interactions with GraphQL using PostgreSQL, achieving a 60% improvement in response times. Led client engagement and requirement gathering, using Figma and Adobe XD for prototypes. Translated feedback into design improvements, boosting user interaction by 40% and client satisfaction by 20%.

Constructed ETL pipelines using Python and Apache Airflow for large-scale data processing, decreasing data latency by 50%. Executed data migrations from legacy to cloud-based solutions (AWS Redshift, Snowflake), which reduced storage costs by 30%. Refined complex SQL queries and indexing strategies, contributing to more efficient data retrieval for analytics and reporting.