The Good Tech Companies - In Conversation With Dung Le: Engineering Excellence Across Tech Giants and Entrepreneurial Ventures

Episode Date: August 21, 2025

This story was originally published on HackerNoon at: https://hackernoon.com/in-conversation-with-dung-le-engineering-excellence-across-tech-giants-and-entrepreneurial-ventures. Dung Le blends engineering excellence and entrepreneurship, from Google and Tesla to startups, driving innovation in performance, data, and consumer tech. Check more stories related to programming at: https://hackernoon.com/c/programming. You can also check exclusive content about #tech-entrepreneurship, #performance-optimization, #dung-le, #distributed-systems, #data-infrastructure, #consumer-tech-platforms, #competitive-programming, #good-company, and more. This story was written by: @sanya_kapoor. Learn more about this writer by checking @sanya_kapoor's about page, and for more stories, please visit hackernoon.com. Dung Le’s journey spans Google, Tesla, Facebook, Citadel, and entrepreneurial ventures like Licensed To Glow. He optimized Go tooling for millions of developers, built large-scale data systems at Tesla and Facebook, and now leads a consumer-facing beauty tech platform. With competitive programming roots and multi-stack expertise, he exemplifies engineering leadership and startup innovation.

This audio is presented by Hacker Noon, where anyone can learn anything about any technology. In conversation with Dungle, engineering excellence across tech giants and entrepreneurial ventures, by Sanya Kapoor. The landscape of modern software engineering demands a unique combination of technical excellence, entrepreneurial vision, and the ability to scale solutions across diverse industries. From optimizing performance in distributed systems to iterating quickly when building consumer-facing platforms, today's technology leaders must navigate complex technical challenges while maintaining a strict focus on business outcomes. This intersection of engineering prowess and business acumen represents the evolution of a new crops of software startup founders in an increasingly competitive digital marketplace. The most impactful engineering founders often demonstrate their capabilities across multiple domains,

from infrastructure optimization at scale to consumer product development. These professionals understand that technical excellence alone is insufficient, successful software solutions required deep understanding of user needs, market dynamics, and a clear go-to market path. The ability to transition seamlessly between enterprise-scale engineering challenges and entrepreneurial ventures showcases the versatility required in today's technology landscape. With experience spanning some of the world's most prominent technology companies and successful entrepreneurial ventures, Dungle exemplifies this modern approach to software engineering. His journey from competitive programming champion to software engineer at Google, Citadel, Facebook, and Tesla, culminating in his current role as co-founder and CTO of Licensed to Glow, which demonstrates a comprehensive understanding of both technical excellence and business innovation.

Mastering performance optimization at scale, building high-performance systems that serve millions of users requires sophisticated approaches to monitoring, optimization, and infrastructure management. The most effective solutions combine real-time performance. performance tracking with predictive analytics, enabling development teams to identify bottlenecks before they impact user experience. During his tenure at Google, Dungla focused on critical infrastructure for the Go programming language ecosystem. Working on the Go Tools team gave me deep insights into how language tooling impacts developer productivity at scale.

Dungla explains, reflecting on his experience building performance monitoring tools for GoPLS, the Go language server, which is used daily by millions of Go developers. I implemented comprehensive performance monitoring that benchmarks all new language features relative to the baseline, with dashboards that visualize performance fluctuations across the entire development lifecycle. Because Go is the backbone for many critical services for Google Cloud Platform, any fluctuations, even minor ones, can greatly affect Customs Workload running in the cloud. This approach to performance optimization extends beyond traditional metrics tone compass developer experience and productivity. By implementing major IDE features including refactoring tools, diagnostics, and in editor documentation, he directly improved how millions of developers interact with Go. The integration of automated workflows for Go package releases further demonstrate show thoughtful

engineering can streamline complex processes while maintaining reliability and consistency. Scaling data processing and analytics infrastructure. Modern data processing challenges require sophisticated approaches to handle massive volumes of information while maintaining query performance and system reliability. The most successful implementations leverage distributed computing frameworks alongside optimized storage solutions, creating systems capable of processing billions of events with consistent performance characteristics. Dung Lee's experience across multiple technology companies highlights various approaches to these challenges. At Tesla, he built production-ready sampling features for IoT telemetry databases,

processing over a billion events daily while maintaining sub-second query performance. At any given time, approximately 2 million Tesla vehicles operate on U.S. roads and millions of energy devices, home batteries, solar, inverters, and chargers, are online. Telemetry from their sensors and cameras streams into Tesla's data platform, is processed through ingestion pipelines, and lands in short-term, hot, stores and long-term archives. Capturing all signals at full fidelity is cost-prohibitive and slows incident response. What matters most for reliability is the context immediately around anomalies, the minutes before and after a device or subsystem goes offline. To address this, I designed and

implemented an event-aware sampling and tiered retention strategy that preserves full fidelity windows around failures while down-sampling routine telemetry. This approach maintains diagnostic signal, reduces storage, compute overhead, and accelerates root cause analysis, improving fleet uptime, customer experience, and grid-integrated reliability at a national scale, he notes, describing his approach to optimizing Spark SQL queries and Spark jobs in Scala. The technical implementation involves sophisticated data partitioning strategies, using derived hash columns to distribute workloads evenly across Spark executors while implementing predicate push-down optimization to minimize retrieved entries. This multi-layered approach to optimization, combined with careful Spark cluster

parameter tuning, enabled significant performance improvements while maintaining data integrity across massive datasets. At Facebook, similar principles applied to analytics infrastructure development, where Dungla implemented the cube library for UPM analytics libraries. This system generates cube tables that serve as efficient aggregation mechanisms across different levels, supporting reporting and dashboard functionality for large-scale analytics operations. The implementation included production-ready pipelines and APIs for data ingestion into Raptor, Facebook's Columnar storage engine for Presto, alongside compute optimization features that reduce query complexity through biphasic aggregation approaches. Building consumer-facing platforms and experiences,

transitioning from internal enterprise software to consumer products isn't a lift and shift. It demands different choices in architecture, UX, and speed of execution. With the bar for product quality and first impressions higher THA never, winning consumer platforms hide complexity behind a seamless and intuitive interface while delivering enterprise-grade security and speed. Dung Lee's entrepreneurial ventures demonstrate this transition effectively. Asco founder and CTO of Licensed to Glow, he leads the development of a subscription-based platform that enables consumers to book beauty and wellness services on demand. We've built a platform that combines complex logistics optimization with seamless user experience, he explains,

after raising the $1.8 million. The technical challenges involve real-time scheduling and sync across different third-party scheduling software our partners use, dealing with time zone data, safeguarding user P and payment information, all while maintaining the simplicity that consumers expect. The platform architecture leverages modern technologies, including Next, JS for front-end development, supabase for back-end, and AWSEKS and Porter to auto-scale UP and down its Kubernetes cluster. This technology stack enables rapid development cycles while providing the scalability necessary for consumer applications with unpredictable usage patterns. Simultaneously, Dung Lee's role as co-founder and limited partner of Tenten, a modern Vietnamese restaurant

with $500,000 plus revenue, demonstrates how technical skills translate to operational excellence in traditional industries. The comprehensive approach included location scouting, lease negotiation, permitting management, interior design, supplier sourcing, and team building, culminating in the implementation of operational playbooks that optimize efficiency across all restaurant functions. Competitive programming excellence and technical foundation. Exceptional software engineering starts with first principles, decomposing problems, optimizing under constraints, and proving correctness. Competitive programming is a rigorous training ground for exactly these skills,

optimization, mathematical reasoning, and performance-conscious coding that transfers directly to production systems. Dung Lee's achievements in competitive programming demonstrate this foundation comprehensively. His first place finish among 1,745 participants in HackerRanks, hacked the interview two, USA competition, alongside qualifying for the ACMICPC World Finals 2021 and placing 24th among 60 university teams in the USA National Championship, collectively showcases exceptional algorithmic problem-solving capabilities. These accomplishments extend to national and international mathematical competitions, including a silver medal in the Hanoi Open Mathematics Olympiad where he placed 15th among

487 students from eight countries and a silver medal in Asia-Pacific mathematical Olympiad for primary schools, APMOPS, an international math competition organized by Wauchang Institution of Singapore for Australia, Brunei, Cambodia, Vietnam, Indonesia, Malaysia, Singapore, Taiwan, Thailand, and the Philippines. This mathematical foundation provides the analytical thinking necessary for complex system design and optimization challenges encountered in production environments. The translation of competitive programming skills to industry applications appears throughout Dung Lee's career trajectory. The algorithmic thinking developed through competitive programming directly supports the

performance optimization work at Google, the data processing challenge at Tesla and Facebook, and the system architecture decisions required for entrepreneurial ventures. Technology integration and modern development practices. Contemporary software development requires proficiency across diverse technology stacks, from front-end frameworks to back-end systems to distributed systems infrastructure. The most effective engineers demonstrate adaptability across programming languages, architecture design, and development methodologies while maintaining consistency in code quality and system reliability. Dung Lee's technical expertise spans multiple programming languages including TypeScript, JavaScript, C++, Go, Python, and Java, alongside

comprehensive experience with modern frameworks and infrastructure technologies. His toolkit includes TypeScript, Next, JS, product surfaces, Edge APIs, Go, high concurrency services, Python, data tooling, and C, C++ or Java, when low latency or the JVM ecosystem is required, as well as expertise with multiple database technologies, enabling rapid prototyping and production deployment. Equally critical is how those systems ship and run. La standardizes on containerized builds on Kubernetes, practices Canary and Blue Green releases, and defines cloud resources in Terraform so his product scale smoothly without compromising reliability or speed. About Dungla, Dungla is a distinguished software engineer and entrepreneur with extensive experience across

leading technology companies and successful startup ventures. Currently serving as co-founder and CTO of Licensed to Glow, he leads technical development of an innovative beauty and wellness services platform that has raised $1.8 million in funding. His professional journey includes software engineering roles at Google, Facebook, Tesla, Amazon, and Citadel, where has specialized in performance optimization, distributed systems, and analytics infrastructure. With a competitive programming background that includes first place finishes in national competitions and participation in ACMICPC national finals, Dungla combines algorithmic excellence with practical engineering experience. He holds a Bachelor of Science in Computer Science from

the University of Rochester and is currently pursuing a master's in computer information sciences with a focus on cybersecurity at Harrisburg University, while maintaining active involvement in professional organizations including A.C.Mond, the IEE Computer Society. This story was distributed as a release by Echo Spire Media under Hackernoon's business blogging program. Learn more about the program here. Thank you for listening to this Hackernoon story, read by artificial intelligence. Visit hackernoon.com to read, write, learn and publish.