Zach Lichtman · Member of Technical Staff at HATCI

Tools I use

Systems

macOS
Linux (Debian)
Yocto embedded Linux

Languages

Swift
Python
C/C++
Rust
Java
TypeScript
HTML/CSS

AI / ML

Claude & Codex SDKs
PyTorch
Hugging Face (LeRobot/ACT)
MONAI
Kubeflow

AI Safety

Technical Fellow, Wisconsin AI Safety Initiative

Alignment
Interpretability
Evals & Red-Teaming

Cybersecurity

Northwestern CTD

Kali Linux
Penetration Testing
Web Exploitation
Network Security
Auth0
Burp Suite

Frontend & Design

React
Next.js
PatternFly
Tailwind
Vercel
Mapbox
Figma
Adobe CC
Blender
Lucid

Backend & DevOps

Spring Boot
Supabase
Tekton
Argo CD
GitLab
JFrog
Xcode
Git
Jira
Notion

Cloud & Infra

AWS
Docker
OpenShift
Kubernetes
Knative
Ansible
Perses

Spoken

English
Spanish

Case study 01 · Experience · 2026

Hyundai · HATCI

Software Engineering InternInfotainment DevOps & Production Software

HATCI is Hyundai Motor Group's R&D center in the U.S., where Hyundai, Kia, and Genesis engineer what ends up in the car. I work on the tooling that decides whether it is safe to ship.

Scorers: 28
Eval channels: 04
Engineers: 1,000+
Vehicle brands: 03

Research system / 01

VERDICT

Production evaluation harness

Architected VERDICT, the evaluation harness HATCI uses to decide whether a production LLM voice assistant is safe to ship to millions of vehicles. It conducts multi-turn spoken conversations and grades each assistant with 28 scorers across semantic, acoustic, interactional, and stability channels—from deterministic assertions to Prometheus-2, G-Eval, and FLASK judges. Red Hat adopted it into its Automotive Suite for use by other carmakers.

01Multi-turn speech

02Assistant response

03Four-channel scoring

04Ship verdict

Model judgesPrometheus-2G-EvalFLASK

Platform / 02

Developer infrastructure

Building the Red Hat OpenShift platform that standardizes infotainment software development for 1,000+ engineers across Hyundai, Kia, and Genesis's global R&D network, including distributed hardware-in-the-loop testing, Cuttlefish device emulation, Tekton CI, GitOps delivery with Argo CD and Ansible, and embedded Linux build and debug tooling.

Validation / 03

Research partnerships

Partnered with Amazon's voice team and Michigan State University researchers on production integration and pre-release validation of in-vehicle voice features.

Open source

Problems I hit, and engineered away

02 / Systems built from first principles

Topographic ski resort terrain with two routes converging at a meeting point — Route convergence159 Epic + Ikon resorts

PowderMeet

A routing problem · 2026

No existing app could tell two skiers on opposite sides of a mountain where to meet. PowderMeet can.

Watch the live demoRouting and run logging · 1:41Live across 159 Epic and Ikon resorts

Runs Dijkstra over each resort's OpenStreetMap trail-and-lift graph to minimize the slower skier's arrival time, max(tA, tB), under trail topology, lift queues, and ability gaps. α-weighted lift-wait penalties and edge-ID tiebreaks ensure both phones independently converge on the same meeting point. Live across 159 Epic and Ikon resorts.

The server builds an identical graph manifest for every device from Overpass and elevation data, quantized to 15-minute time buckets, 0.5°C, and 10 m elevation for consistent hashing. Geohash-6 Supabase Realtime channels scope presence fan-out by area, while a Postgres state machine with CHECK and trigger guards plus friendship-scoped RLS tracks meet-up requests.

Mapbox
Swift 6
SwiftUI
Supabase (Postgres + Realtime + Edge)

Heartable

A library problem · 2026

A Valentine's mixtape that refused to care which streaming service owned each track.

Built to make a Valentine's mixtape that could span Spotify, Apple Music, and other services: unifies them into one master library behind a provider-adapter layer, where a multi-source player plays each track through the service that owns it, skip-edit song versioning enables mixtape-style playback, and library snapshot/restore round-trips as idempotent CSV for cross-account migration. Also tracks listening stats and a friends leaderboard.

Ground-up SwiftUI rewrite of a React Native app (SPM-only, zero CocoaPods), reusing the existing Supabase backend unchanged: Postgres with RLS, RPCs, edge functions, and server-side OAuth token refresh per connected service.

Swift 6
SwiftUI
Supabase (Postgres + RLS + Edge)
Spotify/Apple Music APIs

A dexterous robotic hand being calibrated on a laboratory workbench — Embodied systemsTeleoperation · imitation learning · navigation

Robotics

Teaching hands to move, and rovers to know where they are

AI/ML Engineer, AlohaMini

WiscoHumanoids · 2026

Teleoperation data collection and imitation-learning policy training (Hugging Face LeRobot, ACT) for the AlohaMini dexterous hand.

Contributed stale observation detection and logging to the LeKiwi client to surface dropped frames during data collection.

Software Engineer, GNSS Navigation

Wisconsin Robotics · 2024-2025

Validated GNSS navigation code for the club's autonomous rover, field-testing across varied terrain and signal conditions.

AI safety

Making models safe, and teaching people to

Safety Scholar

Wisconsin AI Safety Initiative · 2026

Conducts collaborative AI-safety research, backed by awarded model API credits and research resources (20+ papers published org-wide).

AI Enablement & Training

Hyundai America Technical Center (HATCI) · 2026

Authored the team's technical training materials on AI development and AI safety, then co-taught engineers to use the tooling properly: agentic loops, agents, and evaluation practice with safety in mind.

Selected contributions

More work, in context

Consulting, open-source contributions, security research, hackathons, writing, and the communities I help build.

Consulting

Brainsprout & Raden Wellness

2025-2026

AI Engineering Healthcare Consultant

Engineered PyTorch/MONAI pipelines rendering anatomically accurate 3D brain models for neurosurgical education & expert testimony (Brainsprout, led by Dr. Michael Walsh, Head of Neurosurgery at Northwestern Medicine).

Guiding AI feature decisions and HIPAA-compliant data handling for Raden Wellness's health platform.

JRNL-iOS

2026

Contributor on alichtman/jrnl-ios

Contributor to an encrypted, git-backed journaling iOS app. Shipped Time Machine, a per-entry commit-history browser that reconstructs entry state at any commit via libgit2 blob reads, and bridged Clibgit2 push/fetch/merge into reusable Swift Repository extensions.

Built the GitProviderForPAT protocol family, unifying GitHub, GitLab, and self-hosted PAT sign-in behind one Swift interface.

Independent Bug Bounty

2022-2024

Cybersecurity Researcher

Identified and responsibly disclosed an IDOR (insecure direct object reference) vulnerability in OpenAI's ChatGPT.