Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

Hi there! I’m Thuy Tran Le, a Senior Software Engineer focused on building real-time AI-powered platforms and media features. I work with WebRTC, WebSockets, React, Next.js, TypeScript, Node.js, and Python to create scalable, low-latency products that power thousands of concurrent users. I love turning complex media and communication challenges into smooth user experiences. Beyond code, I design end-to-end AI pipelines, integrate voice and avatar rendering using Whisper, Google Speech-to-Text, Tacotron/Coqui TTS, and Stable Diffusion/OpenAI for in-chat generation. I’m passionate about delightful UX—typing animations, session continuity, and multimedia outputs—and I build robust backends (FastAPI, Express) deployed on AWS with Docker, Kubernetes, and CI/CD to deliver reliable, scalable features.…Hi there! I’m Thuy Tran Le, a Senior Software Engineer focused on building real-time AI-powered platforms and media features. I work with WebRTC, WebSockets, React, Next.js, TypeScript, Node.js, and Python to create scalable, low-latency products that power thousands of concurrent users. I love turning complex media and communication challenges into smooth user experiences. Beyond code, I design end-to-end AI pipelines, integrate voice and avatar rendering using Whisper, Google Speech-to-Text, Tacotron/Coqui TTS, and Stable Diffusion/OpenAI for in-chat generation. I’m passionate about delightful UX—typing animations, session continuity, and multimedia outputs—and I build robust backends (FastAPI, Express) deployed on AWS with Docker, Kubernetes, and CI/CD to deliver reliable, scalable features.

Thuy Tran Le

Full Stack Developer, Web Developer, Back-End Developer, +2





Hi there! I’m Thuy Tran Le, a Senior Software Engineer focused on building real-time AI-powered platforms and media features. I work with WebRTC, WebSockets, React, Next.js, TypeScript, Node.js, and Python to create scalable, low-latency products that power thousands of concurrent users. I love turning complex media and communication challenges into smooth user experiences. Beyond code, I design end-to-end AI pipelines, integrate voice and avatar rendering using Whisper, Google Speech-to-Text, Tacotron/Coqui TTS, and Stable Diffusion/OpenAI for in-chat generation. I’m passionate about delightful UX—typing animations, session continuity, and multimedia outputs—and I build robust backends (FastAPI, Express) deployed on AWS with Docker, Kubernetes, and CI/CD to deliver reliable, scalable features.…Hi there! I’m Thuy Tran Le, a Senior Software Engineer focused on building real-time AI-powered platforms and media features. I work with WebRTC, WebSockets, React, Next.js, TypeScript, Node.js, and Python to create scalable, low-latency products that power thousands of concurrent users. I love turning complex media and communication challenges into smooth user experiences. Beyond code, I design end-to-end AI pipelines, integrate voice and avatar rendering using Whisper, Google Speech-to-Text, Tacotron/Coqui TTS, and Stable Diffusion/OpenAI for in-chat generation. I’m passionate about delightful UX—typing animations, session continuity, and multimedia outputs—and I build robust backends (FastAPI, Express) deployed on AWS with Docker, Kubernetes, and CI/CD to deliver reliable, scalable features.

Available to hire

Hi there! I’m Thuy Tran Le, a Senior Software Engineer focused on building real-time AI-powered platforms and media features. I work with WebRTC, WebSockets, React, Next.js, TypeScript, Node.js, and Python to create scalable, low-latency products that power thousands of concurrent users. I love turning complex media and communication challenges into smooth user experiences.

Beyond code, I design end-to-end AI pipelines, integrate voice and avatar rendering using Whisper, Google Speech-to-Text, Tacotron/Coqui TTS, and Stable Diffusion/OpenAI for in-chat generation. I’m passionate about delightful UX—typing animations, session continuity, and multimedia outputs—and I build robust backends (FastAPI, Express) deployed on AWS with Docker, Kubernetes, and CI/CD to deliver reliable, scalable features.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Work Experience

Senior Software Engineer at Cognizant

May 1, 2024 - Present

Initiated a real-time AI video call prototype using WebRTC, RTCPeerConnection, and Socket.IO, reducing prototype time by 60% and validating audio-video sync for 1,000 concurrent test users. Architected a low-latency media pipeline leveraging FFmpeg and GPU inference, lowering end-to-end media latency to under 800ms for avatar rendering. Implemented voice input/output integration with Whisper, Google Speech-to-Text, and Tacotron/Coqui TTS, increasing speech recognition accuracy to 92% and reducing TTS latency by 30%. Built backend orchestration services in FastAPI and Node.js/Express, exposing REST and gRPC endpoints and delivering high-availability for thousands of users. Designed chat UX components in React/Next.js to support typing animations, session continuity, and multimedia outputs, improving retention in experiments.

Senior Full-Stack Engineer at EPAM Systems

July 1, 2021 - April 1, 2024

Scaled real-time chat backend to support 5,000 concurrent WebSocket connections using Node.js, Socket.IO, and horizontal sharding, improving concurrency by 4x. Enhanced avatar synchronization by integrating WebRTC Data Channels and PostgreSQL, increasing latency synchronization across multi-device flows. Spearheaded AI orchestration layers in Python with FastAPI to route requests to OpenAI GPT and on-prem TensorFlow services, increasing throughput to 2,000 requests per minute. Migrated monolithic services to microservices with Docker, Kubernetes, and Helm, shortening deployment lead time from days to hours and improving fault isolation by 70%. Tuned database performance through PostgreSQL indexing, query optimization, and Redis caching, reducing average session lookup latency by 47%. Validated voice synthesis quality with AB testing across Tacotron and Coqui, selecting a pipeline that raises user satisfaction by 18% in pilot tests. Deployed CI/CD pipelines using GitHub Actions and Terr

Senior Full-Stack Engineer at Netguru

February 1, 2018 - June 1, 2021

Reduced page load time by 33% using server-side rendering in Next.js, code-splitting, and TypeScript optimizations across chat and media UIs. Increased image generation throughput by integrating Stable Diffusion GPU workers, processing 120 images per GPU-hour and improving response consistency by 22%. Coordinated cross-functional agile sprints with product and design to define typing animations, multimedia cards, session UX, driving feature adoption. Mentored 6 junior engineers on real-time patterns, WebRTC, and media pipelines, boosting team delivery velocity by 15% within 6 months. Launched containerized deployments with Docker and Kubernetes, automating release pipelines and reducing deployment lead time. Automated media servers with ELK Stack and monitoring, improving incident triage time by 40% during testing and demos. Experimented with on-device inference using TensorFlow Lite, reducing network calls by 40% for offline features. Prototype avatar rendering flow with Three.js and

Software Engineer Intern at FPS Software

April 1, 2017 - January 1, 2018

Configured CI pipelines in Jenkins and Git for Node.js and Python, cutting build failures by 27%. Streamlined container builds and developer workflows using Docker Compose, reducing environment setup time from hours to 15 minutes. Synthesized model deployment notes for PyTorch and TensorFlow, advising on patterns to deploy models with under 2 GB memory footprint. Enabled database migrations and automated backups for PostgreSQL, creating scripts that reduce restore time to under 10 minutes. Explored speech-to-text solutions including Whisper and cloud STT, benchmarking latency and accuracy across five datasets. Launched lightweight chat prototype using React, Express, and Socket.IO.