Skip to content
View tmacraft's full-sized avatar

Block or report tmacraft

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
tmacraft/README.md

Hi, I'm Lu Wang 👋

AI engineering leader based in New York. I build and scale machine learning systems for search, personalization, and conversational AI, and I stay hands-on with the technology while leading large teams.

Background

  • 10+ years building ML products at Fortune 500 companies: Verizon, Restaurant Brands International (Burger King, Popeyes, Tim Hortons), and Office Depot.
  • Currently leading a global AI science team at Verizon, moving a legacy chat and search platform to an agentic, AI-native virtual assistant.
  • Earlier work covers recommendation engines, real-time personalization, and a voice ordering platform.
  • Lead author on a CIKM 2023 paper on multi-task recommenders.

What I'm working on now

  • Agentic AI: multi-agent orchestration and production guardrails, built on Google's Agent Development Kit and Gemini.
  • Recommender systems: multi-task and Transformer-based models for search and recommendation at scale.
  • Core CS foundations: systems programming, algorithms, and the math behind machine learning, through graduate coursework.

Tools and technologies

Python Java C Apache Spark Ray TensorFlow BigDL Google ADK GCP AWS Kubernetes

Problems I like

  • Getting machine learning out of notebooks and into reliable, measurable production systems.
  • Personalization and recommendation when latency, scale, and business outcomes all matter.
  • Making generative and agentic AI safe, explainable, and testable.

Goals

  • Keep building depth in agentic systems and distributed ML infrastructure.
  • Strengthen my computer science fundamentals and put them into public, well-documented projects here.
  • Contribute more to open-source AI frameworks.

Selected work

  • Paper: "Efficient Multi-Task Learning via Generalist Recommender," Proceedings of the 32nd ACM International Conference on Information and Knowledge Management (CIKM 2023, lead author)
  • Talks: Ray Summit 2021 and 2023, on large-scale recommendation with Ray and Spark
  • Open source: six contributions landed in Google's Agent Development Kit (adk-python), including automatic session creation, a Google Search tool model option, and fixes to skill loading, state handling on resume, and event rehydration performance. These are under my work account, @lwangverizon. I also contributed to Intel's BigDL (see my forks below).

Connect

LinkedIn

Pinned Loading

  1. analytics-zoo analytics-zoo Public

    Forked from intel/BigDL

    Distributed Tensorflow, Keras and BigDL on Apache Spark

    Jupyter Notebook

  2. ray ray Public

    Forked from ray-project/ray

    A fast and simple framework for building and running distributed applications. Ray is packaged with RLlib, a scalable reinforcement learning library, and Tune, a scalable hyperparameter tuning libr…

    Python

  3. tmacraft tmacraft Public

    Profile README