About

Engineer between models and people.

I'm Dingxin Tao — 陶鼎新 — an AI full-stack engineer and forward deployed engineer based in Southern California.

Dingxin Tao

The story

I got into software because I liked making things that other people could actually use. That instinct pulled me toward the full stack early: the interface people touch, the API underneath it, and the data that makes it honest.

Then large language models changed what 'usable' could mean. I spent the following years learning how to make AI features that survive production — retrieval that stays grounded, agents that stay on task, and evaluation loops that tell you the truth before your users do.

Forward deployed engineering is where those threads meet. I embed with customers, learn their workflow, integrate with their systems, and iterate until the solution works in their world. Then I bring what I learned back to the product.

I'm finishing my B.S. in Computer Science at UC Irvine in 2026 and continuing into the M.S. in Analytics at USC for 2027 — because the models are only as good as the data and the decisions around them.

Education

Two campuses, one trajectory.

Class of 2027

2026 — 2027

University of Southern California

Master of Science · Analytics

  • Statistical modeling & forecasting
  • Machine learning for decision-making
  • Data engineering & experimentation
  • Los Angeles, CA
Class of 2026

2022 — 2026

University of California, Irvine

Bachelor of Science · Computer Science

  • Systems, algorithms & software design
  • Machine learning & AI
  • Human-computer interaction
  • Irvine, CA

Experience

Where the work happened.

  1. 2025 — Present

    AI platform · enterprise deployments

    Forward Deployed Engineer

    Embedded with enterprise customers to take AI products from pilot to production inside their environments.

    • Led discovery, solution design and rollout for RAG and agent deployments in regulated environments.
    • Built integrations against customer identity, document and ticketing systems.
    • Owned evaluation harnesses that gated every prompt and model change.
    • Python
    • TypeScript
    • LangGraph
    • PostgreSQL
    • AWS
    • Docker

  2. 2023 — 2025

    Product teams · startups & labs

    AI Full-Stack Engineer

    Designed and shipped LLM-powered product features end to end — from data model to streaming UI.

    • Shipped agentic workflow tooling with typed tool schemas and replayable runs.
    • Built observability for model calls: traces, costs and model-graded quality.
    • Designed component systems and motion for React / Next.js frontends.
    • Next.js
    • React
    • FastAPI
    • ClickHouse
    • Redis
    • Vercel

  3. 2022 — 2023

    Internships & university projects

    Software Engineer (Full-Stack)

    Learned the full stack by shipping: web apps, APIs, databases and the deployments that kept them online.

    • Built data-driven web applications with React and Node.
    • Set up CI/CD, containerization and monitoring for student and client projects.
    • React
    • Node.js
    • PostgreSQL
    • Docker
    • GitHub Actions

Now

Currently

  • 01Finishing my B.S. in Computer Science at UC Irvine (Class of 2026).
  • 02Preparing for the M.S. in Analytics at USC (Class of 2027).
  • 03Building agentic systems with typed tool-calling and evaluation harnesses.
  • 04Open to forward deployed engineering and AI engineering roles.
Say hello

Values

01

Ownership

From discovery call to production dashboard — if it's mine, it's mine end to end.

02

Clarity

Plain language for stakeholders, typed contracts for machines.

03

Momentum

Small, shipped increments beat perfect plans. Observe, then improve.

04

Craft

The details people feel but can't name — latency, motion, empty states — matter.