Hi, my name is
Dhruv Bhanderi
I build scalable systems & intelligent platforms.
Software Engineer specializing in building scalable full-stack applications, cloud infrastructure, and high-performance systems. Currently pursuing MS in Computer Science at USC.
01.
About Me
I'm a Software Engineer pursuing my Master's in Computer Science at USC (3.85 GPA), with a B.Tech in CS from Pandit Deendayal Energy University (3.88 GPA). I specialize in building production-grade distributed systems, scalable APIs, and AI-powered platforms.
I've contributed to Kubernetes (v1.36), built LLM diagnostic agents with Claude API, engineered Raft-based distributed file systems, and deployed GenAI pipelines processing 1.5M+ tokens for World Bank and Gates Foundation research. My stack spans Go, Python, TypeScript, Django, Kafka, gRPC, AWS, Docker, Kubernetes, Helm, and Terraform.
Currently working at the USC Annenberg Norman Lear Center building NLP pipelines and semantic retrieval systems, with 2 IEEE publications and 8+ merged open-source PRs across Kubernetes, Glean, OWASP, and pnpm.
Education
University of Southern California
Master of Science in Computer Science (GPA: 3.85/4)
Sep 2024 – May 2026
Pandit Deendayal Energy University
Bachelor of Technology in Computer Science & Engineering (GPA: 3.88/4)
Nov 2020 – May 2024
Currently
Working at
USC Norman Lear Center
Studying
NLP & Information Retrieval @ USC
Building
AI-powered platforms & distributed systems
Based in
Los Angeles, CA
02.
Skills & Technologies
A comprehensive toolkit built through real-world projects and production systems. Hover over each category to explore.
Languages
Frontend
Backend
Cloud & DevOps
Databases & Messaging
ML / NLP / AI
By the numbers
Impact in Numbers
Years of Experience
Projects Built
Open Source PRs Merged
Technologies
IEEE Publications
Certifications
03.
Work Experience
Building production systems at scale. Click on each role to see details.
Software Engineer
USC Annenberg Norman Lear Center
- Architected a Django + PostgreSQL backend on AWS Elastic Beanstalk; optimized ORM queries, connection pooling, and B-tree indexing to boost API throughput 42% and reduce p95 latency 35% under peak load.
- Built scalable ETL pipelines processing 50K+ multilingual data points and 1.5M+ tokens into PostgreSQL + pgvector, with async I/O throttling reducing API rate-limit failures to <2%.
- Built GenAI-native data pipelines leveraging LLM classification (OpenAI, Claude) for sentiment/toxicity analysis on multilingual social media data for World Bank and Gates Foundation research.
- Engineered a semantic retrieval system using text-embedding-3-large with pgvector for 150K+ records; evaluated inter-annotator agreement (Cohen's kappa = 0.576, Krippendorff's alpha).
- Configured CI/CD pipelines with rollback support via GitHub Actions; automated Docker image builds and health-probe validation, increasing release reliability 40%.
Software Development Engineer
SPIGOT Developers
- Designed Node.js/Express microservices with Redis caching and Kafka message queues, improving p95 latency 35% and doubling throughput (2.1x) across 14+ stable releases via horizontal scaling and async I/O.
- Streamlined CI/CD via GitHub Actions with parallel tests and cached Docker images, cutting build-to-deploy time from 18 min to 7 min (-61%) and enabling one-click rollbacks.
- Refactored a monolithic codebase into modular microservices with gRPC APIs, distributed tracing, and rate limiting, accelerating deployment velocity 3x across environments.
Full Stack Developer Intern
SPIGOT Developers
- Built modular Vue.js + TypeScript + Tailwind CSS dashboards serving 300+ DAUs with responsive, accessible UX.
- Implemented OAuth2 + RBAC authorization, blocking 90% of unauthorized requests; deployed serverless functions for real-time analytics.
- Wrote 50+ unit and integration tests (Jest, Cypress), achieving 85% code coverage; resolved 12 WCAG accessibility violations across 6 modules.
04.
Featured Projects
A selection of projects I've built. Filter by technology to explore.
DeployIQ — AI K8s Deployment Pipeline
Production-grade Go CLI for Kubernetes deployments with an LLM diagnostic agent using Claude API with native function calling. Iteratively invokes 5 K8s tools across up to 5 reasoning iterations. RAG pipeline with pgvector for incident retrieval. Kafka event streaming, 6 custom Prometheus metrics, Terraform IaC, and <30s automated rollbacks.
Distributed File Storage System
GFS-inspired distributed file system built from scratch with Raft-replicated master cluster for leader election, log replication, and automatic failover. 16-shard metadata partitioning with chain-replication writes across N-way replicas, benchmarked at 47K+ ops/sec (64 concurrent clients, p99 latency 2.3ms).
Reddit:AI — Intelligence Platform
AI-powered platform built at CalHacks 12.0 with real-time sentiment and toxicity classification using Claude. Integrated Letta AI agents with persistent memory, reducing redundant moderation flags 80% vs. AutoMod. Built with Next.js, FastAPI, and MongoDB.
05.
Publications
Peer-reviewed research published at IEEE conferences.
Bridging Dialects: Enhancing Neural Machine Dialect Translation for Gujarati Language Through Model Adaptation
IEEE INDIACom 2025
Thakkar, Vekariya, Bhanderi, Paliwal
Impact of Two-Factor Authentication on User Convenience and Security
IEEE INDIACom 2023
Bhanderi et al.
06.
Open Source Contributions
Merged pull requests across major open-source projects.
Kubernetes
kubernetes/kubernetes #137189
Reviewed by @liggitt (senior K8s maintainer)
+92, -3
Fixed client-go fake Pods.GetLogs to honor test reactors in v1.36. Made GetLogs reactor-aware, propagated reactor errors, and wrote comprehensive unit tests.
Glean Agent Toolkit
gleanwork/glean-agent-toolkit #25
Fixed Pydantic serialization so camelCase alias fields survived ADK adapters. Added regression tests for SDK compatibility.
pnpm
pnpm/pnpm #10711
Fixed cached bundled-manifest behavior so devDependencies persist in stored manifest subsets.
OWASP BLT
OWASP-BLT/BLT #5948
Fixed N+1 query in the issues API with select_related, prefetch_related, and annotated counts.
Glean MCP Server
gleanwork/mcp-server #342
Added GHCR visibility validation and runtime smoke testing for Dockerized MCP server releases.
05.
Certifications & Achievements
Professional certifications that validate my expertise.
IBM Full Stack Software Developer
Proficiency in cloud-native application development with React, Node.js, Docker, and Kubernetes.
Foundations: Data, Data, Everywhere
Foundational data concepts and their application in data-driven decision-making.
IBM Full-Stack JavaScript Developer
Full-stack JavaScript development with modern frameworks and cloud deployment.
06.
Get in Touch
I'm currently open to new opportunities. Whether you have a question or just want to say hi, my inbox is always open.