KeploreAI Inc
Software Engineer — AI Systems
Developing production LLM inference workflows, model-serving microservices, telemetry, and REST infrastructure for AI-powered enterprise applications.
01Machine Learning Systems Engineer
I build production AI inference services and distributed backend systems, with a focus on performance, observability, and system reliability.
Résumé at a glance
02Experience
From GPU inference optimization to distributed APIs, my work connects model performance with production reliability.
KeploreAI Inc
Developing production LLM inference workflows, model-serving microservices, telemetry, and REST infrastructure for AI-powered enterprise applications.
Amazon
Optimized a production robotics perception pipeline with TensorFlow and TensorRT, improving OCR accuracy from 86% to 90% and reducing latency from 300 ms to 50 ms.
University of Houston
Built data processing and indexing systems for large-scale video datasets, developed video-analysis and mesh-evaluation prototypes, and taught data-systems courses.
FloAuto
Refactored a legacy ERP into modular C++ microservices, optimized multithreaded data pipelines from days to hours, and implemented Elasticsearch indexing and monitoring.
03Education
M.S.E. in Electrical & Computer Engineering
Baltimore, MDB.S. in Computer Science · Minor in Mathematics
Cum Laude · Dean's List, 6 semesters04Technical skills
Python · C++ · Java · Kotlin · TypeScript
LLM Infrastructure · Model Serving · RAG · Prompt Engineering · Evaluation · Guardrails
LangChain · LangGraph · MCP · Graph Retrieval · Embeddings · Human-in-the-loop
TensorFlow · PyTorch · TensorRT · GPU Inference · Computer Vision · OCR
Distributed Systems · Linux · Microservices · REST APIs · Docker · Cloudflare Workers
SQL · Redis · MongoDB · SQLite · CI/CD · Testing · Telemetry
05Research & publications
Published work spanning mesh-structure visualization and fine-grained facial emotion recognition.
Visualization workflows and structural analysis for high-fidelity hexahedral mesh evaluation.
Si, Lei; Cao, Haowei; Chen, Guoning · IEEE Xplore ↗An end-to-end deep learning pipeline for fine-grained facial emotion recognition and model evaluation.
Powers, Gregory; Jin, Aobo; Tonmoy, Abdul Basit; Cao, Haowei; Gohel, Hardik; Deng, Qixin · DOI ↗These are selected highlights, not the full boundary of the work. Additional collaborations are moving forward, including behind-the-scenes engineering and resource support for adjacent research.
07Life outside the terminal
A Rottweiler puppy with serious eyebrows, excellent debugging presence, and a growing collection of naps.
Born on February 15, 2026. She keeps the site human, reminds me to step away from the screen, and now helps visitors explore my résumé.





08Let's talk
I’m interested in machine learning systems, AI infrastructure, and backend roles where performance and reliability matter.
For recruiting, collaboration, or technical conversations.
Complete a quick human check before continuing to the scheduling page.
Continue to Calendly↗