UMANG GOEL
Backend & Database Engineer with hands-on experience building scalable services and optimizing data workflows. Delivered a multi-mode SRT retention system reclaiming ~2GB (~84%) storage and cutting query latency by 35%, and built enterprise REST APIs and Airflow ETL pipelines processing 50K+ records/run with 25% lower latency. Strong in Java, Python, SQL, Docker, AWS, and CI/CD.
TECHNICAL SKILLS
Arizona State University · B.S. Computer Science · Graduating May 2026
I'm a Computer Science student at Arizona State University with hands-on experience building scalable backend services and optimizing complex data workflows. Experienced in architecting enterprise REST APIs, stored procedures with REGEXP input validation, and automated Airflow ETL pipelines.
Demonstrated ability to improve system latency and database storage efficiency (reclaiming ~2GB / ~84% storage and cutting p95 query latency by 35%). Graduating May 2026 and focused on testing, automation, and reliable backend engineering.
Arizona State University
Aug 2022 – May 2026
Bachelor's, Computer Science
Tempe, AZ
Coursework
Core Concepts
EXPERIENCE
DigiClips Media Analytics Platform
Backend & Database Engineer Intern (Capstone)
Tempe, AZ
-
Designed 6 MySQL stored procedures with REGEXP-based input validation (email, phone, ZIP) and centralized error logging, reducing invalid data ingestion errors by ~40% across 3 backend services.
-
Built a multi-mode SRT retention system with preview, execution, and reporting modes; enforced a 90-day cleanup policy, reclaiming ~2GB (~84%) of total database storage.
-
Optimized SRT ingestion and clip metadata queries using composite indexing and query rewrites, reducing p95 latency by ~35% under sustained load.
-
Deployed a Dockerized MySQL 8.x instance over OpenVPN; resolved production issues (duplicate procedure conflicts, DNS misconfigurations), maintaining 99%+ uptime over a 4-month sprint cycle.
Honeywell and ASU Innovation Hub
Software Developer Extern
Tempe, AZ
-
Developed and maintained enterprise RESTful services in Java and Python, implementing scalable API design and optimizing request handling to reduce end-to-end latency by 25%.
-
Built Airflow ETL pipelines processing 50K+ records per pipeline execution using PostgreSQL and Redis, enabling reliable analytics ingestion.
-
Wrote 100+ unit and integration tests (xUnit, pytest), resolving 30+ defects with QA and reducing post-release issues by 10%.
DMML Lab, Arizona State University
Software Developer Intern
Tempe, AZ
-
Developed distributed backend services using Python and Java, applying object-oriented design patterns and Agile practices to reduce p95 latency by 30%.
-
Built event-driven pipelines with RabbitMQ handling 500+ events/min; designed data workflows processing 300K+ records with validation and concurrency controls, improving analytical accuracy by 15%.
-
Implemented TDD (75%+ coverage), CI/CD pipelines with Docker, and documentation standards, reducing production defects by 40% across 3 service modules.
PROJECTS
High-performance systems and data-intensive applications
Real-Time Event Processing Engine
Designed and implemented a high-throughput event processing system using Python services and Redis Streams, processing 10K+ events/sec across 4 worker nodes with consistent sub-50ms p95 latency.
- ▸ Implemented retry logic with 3-tier escalation, fault handling, and 60+ automated test cases ensuring 99.9% reliable event processing under peak load.
- ▸ Monitored and triaged 15+ production incidents using structured logging & dashboards, identifying root causes within 10 min and cutting recurring failures by 50%.
Causal MMD
Designed scalable Flask application integrating 1.03M+ records with ensemble classifiers, achieving 92% accuracy for real-time disaster response coordination.
- ▸ Applied performance optimization through async I/O and batch processing with connection pooling, scaling throughput to 10K+ points/min and cutting latency by 30%.
- ▸ Built thread-safe REST API with versioned endpoints and OpenAPI docs, containerized with Docker and deployed via CI/CD with 95%+ test coverage.
PUBLICATION
An Interventional Approach to Real-Time Disaster Assessment via Causal Attribution
Saketh Vishnubhatla, Alimohammad Beigi, Rui Heng Foo, Umang Goel, Ujun Jeong, Bohan Jiang, Adrienne Raglin, Huan Liu
ACM Digital Library — dl.acm.orgCURRENTLY
High-Perf Networking Library
Async I/O event loop in C++ using io_uring for zero-copy networking, targeting sub-microsecond latency for financial data feeds.
Distributed Storage Internals
Deep-diving into LSM-tree design, compaction strategies, and write-ahead logging — inspired by RocksDB and LevelDB internals.
Systems Programming Papers
Working through "Designing Data-Intensive Applications" and recent VLDB/OSDI papers on distributed query execution and storage engines.
CONTACT
Graduating May 2026 · Actively seeking full-time Backend & Database / Software Engineering roles.
Passionate about building fast, scalable backend systems and optimizing data pipelines.