Language Proficiency
C/C++ 95%I'm a Computer Science student at the University of California, Los Angeles (UCLA), expected to graduate in June 2026. My academic focus revolves around backend systems, cloud infrastructure, and AI/ML. I'm exploring advanced approaches to distributed systems and scalable architectures through coursework in Distributed Systems, Machine Learning, and Natural Language Processing.
My development toolkit includes Python, Go, Java, TypeScript, C++, React, Next.js, Node.js, Kubernetes, Docker, AWS, and GCP. Through SWE internships at Verkada and Square, along with research at UCLA's NLP Lab where I published on bias control in LLMs, I've built expertise in cloud infrastructure, distributed systems, and AI/ML.
Through a variety of software engineering experiences across cloud infrastructure and backend systems, I've honed my ability to architect and build reliable, scalable systems. My vision is to build and lead technology companies that solve real-world problems. I'm excited to continue applying my passion for backend cloud design, push the boundaries of what's possible in systems engineering, and eventually start companies that make an impact at scale.
Experiences
Technical Software Experiences
Verkada
Software Engineering Intern - Data Infrastructure
September 2025 - December 2025
Architected deployment of automated EKS Resource Optimizer across 500+ services, automating daily analysis of 30 days of Datadog metrics and replacing manual tuning with fully self-serve pipeline. Delivered $XXX,XXX+ in annualized AWS cost savings by optimizing 1000+ workloads, reducing CPU requests by 68% and memory requests by 79% while maintaining zero production-impacting incidents. Led cross-team pilot and phased rollout achieving 98%+ retention and org-wide adoption within one quarter.
Square
Software Engineering Intern - Payment Balances
June 2025 - September 2025
Spearheaded migration of Financial Accounts Admin from legacy GCP Books to AWS Aurora MySQL, modernizing payments infrastructure supporting 4.5M accounts and processing 2.5M+ daily transactions. Migrated GraphQL and Elastic Graph services from multi-AZ to single-AZ architecture with Terraform, eliminating cross-zone network overhead and saving $XX,XXX annually. Built Java gRPC services and TypeScript frontend to enable bulk reimbursements and SEV-related outage adjustments, reducing monitoring time from 78 minutes to 13 minutes per bulk outage.
FocusAI (Acquired by Qualcomm)
Software Engineering Intern
June 2024 - September 2024
About the Company:
FocusAI delivered Video Surveillance as a Service (VSaaS) solutions using AI for real-time analytics. I developed scalable AWS infrastructure using Go and Kubernetes, reducing database redundancies by 50% and increasing throughput by 10%. Additionally, I built a comprehensive front-end CI/CD pipeline with automated regression testing (JavaScript, Selenium WebDriver), improving platform stability by catching 95% of failed commits.
Projects
Software-Related Side Projects
CourseBidder
Founder & Full-Stack Developer
June 2024 - June 2025
Project Scope:
CourseBidder LLC is a platform that facilitates the anonymous buying and selling of class seats and other student-to-student services, addressing critical needs for accessibility in course registration.
What I Did:
I founded CourseBidder and developed a full-stack web application using JavaScript, React.js, MongoDB, Stripe, Google OAuth, and Node.js. I wrote over 3000 lines of code to ensure seamless transactions and user experiences. Additionally, I drafted business plans, legal documents, and operational strategies to publicly launch the platform in October 2024, achieving a projected market reach of 3000+ students at UCLA within the first year.
Project Link: Private until public launch
Site: coursebidder.com
Beta Ball Site
Full-Stack Developer
January 2024 - Present
Project Scope:
The Beta Ball Website is a sports league management platform built using HTML/Bootstrap for the front-end, Flask for the backend, and SQLite with SQLAlchemy for database handling.
What I Did:
I developed the platform with features like a Home page displaying key stats, a Stats page with detailed player and team information, a New Season page for setting up teams and players, and a Game page for updating match stats. I also built the Archive page to store past season stats and player cards for displaying individual stats.
Project Link: github.com/atijmahesh/beta-ball
Site: beta-ball.up.railway.app
UCLA NLP Research Lab
Applied AI Researcher
April 2025 - October 2025
Project Scope:
Research focused on compositional bias control in Large Language Models, developing evaluation frameworks and benchmarking control strategies to identify effective alignment approaches.
What I Did:
Developed 7-metric evaluation framework spanning 20 occupations and over 72,000 generated completions across fluency, compliance, diversity, and variance. Published "Compositional Bias Control in Large Language Models: Preference Learning Fails, Supervision Succeeds" (arXiv, 2025), benchmarking 6 control strategies including supervised fine-tuning, preference learning, and linear projection to identify limits of preference-based alignment.
Paper: arXiv, 2025
Countries I've Visited
An interactive map showing the countries I've explored