# Xianglong Hu

> Leader | LLM Researcher | Builder

Location: San Francisco, California, United States
Profile: https://flows.cv/xianglong

For those who are interested, please take a look at my resume, 
https://github.com/hu-xianglong/Resume/blob/master/Resume%20(1).pdf

I have done research in the field of both computer science and physics before,
so I have rich experience in solving problems. I am a programmer with enthusiasm
for sharing and learning.

Currently, I am developing an open source project, a
Microsoft Language Server for Wolfram Language in Wolfram itself. For details
and other wonderful
projects, please check out my Github(https://github.com/huxianglong). The alpha version VS code client has been
released on VS marketplace as well. This project is really interesting because we
are bootstrapping at the moment, i.e., we're developing it with itself.

I have tried to explored topics as vast as I can in computer science, especially
those fields that are demanding and mathematically challenging. The experience I
have covers fields of machine learning, geometric modeling, compiler construction, and
distributed systems, etc. 

I am seeking positions in software development.

## Work Experience
### Open Source Developer @ Open Source
Jan 2025 – Present
Nano-vLLM Architecture Profiler (Open Source Project) | Feb 2026

Systems Analysis: Conducted a deep-dive architectural analysis of the Nano-vLLM (1.2k LOC) inference engine, mapping the lifecycle of PagedAttention and KV-cache block management.

Simulated Performance Benchmarking: Engineered a Mock Execution Engine to simulate GPU kernel latency, enabling the testing of scheduling algorithms and memory fragmentation on CPU-only environments.

Optimization Research: Identified potential throughput bottlenecks in the First-Come-First-Served (FCFS) scheduler during long-form reasoning (CoT) generation; documented architectural improvements for multi-agent orchestration.

Tooling: Developed a Python-based visualizer for tracking KV-Cache block allocation and reference counting in real-time.

### Software Development Enginner II @ Amazon Web Services (AWS)
Jan 2024 – Present | San Francisco, CA
- Scalable Workflow Orchestration: Led the architectural refactoring of the Outpost infrastructure provisioning service, transitioning to a distributed workflow model using AWS Step Functions and Lambda. This modernization improved system scalability and significantly reduced maintenance overhead through decoupled, stateful execution.

- Lifecycle Automation: Spearheaded the end-to-end design and development of the Outpost Decommission Service, automating complex hardware retirement protocols and ensuring secure, compliant infrastructure turnover.

- Edge Reliability: Maintained and optimized the Snow OTA (Over-the-Air) Update Service, ensuring high-availability and seamless firmware deployments across globally distributed edge computing devices.

### Open Source Developer And Reseacher @ CAMEL-AI.org
Jan 2024 – Present
Scaled LLM Reasoning (Loong Project): Engineered an open-source framework for synthesizing and verifying long-form Chain-of-Thought (CoT) data at scale. Successfully applied Reinforcement Learning on logic data to reproduce the state-of-the-art reasoning results of DeepSeek-R1 and Logic-RL, inducing emergent behaviors like self-reflection and verification.

Reinforcement Learning (RL) Involvement: Involved in the development of the ReaL-TG framework, which utilizes RL to optimize language models for explainable link forecasting on temporal graphs. Contributed to designing reward signals that prioritize transparency and logical consistency in model predictions.

Supervised Fine-Tuning (SFT): Leveraged SFT workflows to distill tool-use knowledge into compact models via back-translated traces, enabling high-performance autonomous agent capabilities with significantly reduced inference overhead.

Temporal Planning & Benchmarking: Designed and published the TCP Benchmark, a specialized evaluation suite for measuring LLM performance on temporal constraint-based planning, bridging a critical gap in multi-step reasoning assessment.

Multi-Agent Orchestration: Contributed to the CAMEL-AI open-source ecosystem, focusing on autonomous communication protocols and the deployment of "societies" of LLM agents to solve complex, distributed tasks.

### Software Engineer @ CloseFactor
Jan 2023 – Jan 2024 | Jersey City, NJ
◦ LLM Product E2E Delivery: Led and developed Account Plans from concept to successful implementation,
utilizing cutting-edge LLM techniques such as openai, rerank, RAG, model fine-tuning. Architected a robust system
incorporating unreliable components like scraper services.
◦ Agile & Lean Startup: Conducted customer interviews and rapidly iterated products based on feedback.
◦ Product Collaboration: Collaborated closely with Product Management to ensure technical and product
alignment.
◦ Impact: Spearheaded the development of our most critical product, contributing to nearly 100% of new ARR.
Received over 30 positive customer reviews on G2. (previously known as deep dive)

### Software Engineer @ DoorDash
Jan 2022 – Jan 2023
◦ Designed and Implemented an in-memory cache for merchant metadata to reduce the Redis traffic.
Meanwhile, refactored the relevant data retrieval path to be more maintainable. Reduced AWS Redis costby 90% and AWS ECR Cost by 50%. Reduced the latency of all tier-1 APIs by 50%.
◦ Reduced the unit tests run time by 60% (15 min to 4 min).

### Software Engineer @ Amazon Web Services (AWS)
Jan 2020 – Jan 2022 | Seattle, WA
◦ Designed and implemented a distributed stateless event processing microservice. Participated in the full cycle of the product feature launch from scoping, designing, implementing, testing and releasing.
◦ Built and tested a new customer-facing API which emits metrics for the customer periodically. Technical stack included AWS EC2, Elastic Search, Dynamo, CloudWatch and SQS.
◦ Took oncall duties bimonthly. Identified, root caused and fixed DevOps issues. Communicated, evaluated and resolved customer tickets.
◦ Made improvements on the oncall experience. Automated all the oncall SOPs with python scripts.
Reduced the latency of one customer-facing API by 90%. Implemented a mechanism controlling the throughput of the system with a dynamic config.
◦ Contributed to the documentations. Maintaining onboarding wikis for the new hires and the oncall

### Software Engineer Internship @ Onai
Jan 2019 – Jan 2019 | Greater New York City Area
• Blockchain & Accumulator: Benchmarked a batching accumulator. Implemented and designed a Merkle-tree like membership-proof
with Celo pairing groups in Rust.
• Git as NoSql, Rust: Designed a key-value store with git to achieve a version-controlled nosql database with minimal dependencies. Utilizing
git internal data structures like blob, tree, and reference to store objects while using path as key. Implemented and tested in Rust with git2.
• Idris: Designed an LL(0) domain-specific language for state machines, which is compiled into Idris to reduced repetitious skeleton code and
javascript to visualize a state machine. The grammar got verified in Antlr. Implemented a VS code frontend for the backend auto-generating
Idris code for Idris holes.

### Research Assistant @ Experiment Center for Physics, Fudan University
Jan 2018 – Jan 2018 | Shanghai City, China
Designed and implemented an apparatus for acoustic wave generation and experiment.
• Functionality: Measuring and analyzing acoustic properties onAndroid smartphones for pedagogical purposes, primarily Fourier trans-
formation of acoustic waves, real-time data processing and rendering.
• Techniques: Designed the frontend, which provided interactive charts and experiments parameter settings, and backend in Android.
Interacted with hardware by thread programming. Data export supported.
• Publishment: Research results published in Physics Experiments, funded by the National Natural Science Foundation (China).

### Research Assistant @ Laboratory of Computer Vision and Machine Learning, Fudan University
Jan 2017 – Jan 2017 | Shanghai City, China
• Purpose: Mathematically formulated a hierarchical Bayesian neural network and adapted it to incremental learning to eliminate catastrophic forgetting when new classes appeared. Further simplified with variational inference to model true distribution.
• Experiment: Network of Gaussian probabilistic weights implemented with local reparametrization and practical Monte Carlo in Pytorch
to expedite computation. Setting up a remote Linux server from scratch. Manage environments though Anaconda. Automated training
based on scripts ran on servers with Cuda GPUs.

### Research Assistant @ Atomic and Optical Physics Lab， Fudan University
Jan 2015 – Jan 2017 | Shanghai City, China
• Adapted various optimizations including genetic algorithm and Powell’s method to the reconstruction of quantum waves from nonlinear noises in MATLAB.


## Education
### Master of Science - MS in Computer Science
New York University

### Exchange Student in Physics
University of California, Berkeley

### Bachelor in Physics
Fudan University


## Contact & Social
- LinkedIn: https://linkedin.com/in/xianglong-hu-354964153

---
Source: https://flows.cv/xianglong
JSON Resume: https://flows.cv/xianglong/resume.json
Last updated: 2026-04-05