AI / ML Engineer · Dr.-Ing.

Anthony Ngo

Builds systems that perceive, reason, and act, from autonomous driving perception to agentic AI in capital markets.

agent · about Anthony idle
plan retrieve respond
about

Research-grade methods, production systems.

Anthony is an AI/ML engineer with a doctorate in engineering and roughly eight years building perception and simulation systems at scale. His background sits where classical computer vision, sensor modeling, and modern generative AI meet.

Today he works in the CTO area at Deutsche Börse, designing agentic AI systems for capital-markets infrastructure. Before that, years at Robert Bosch on autonomous driving simulation: synthetic data, Radar/Lidar sensor models, perception, and neural rendering. He likes problems where research methods have to survive contact with real systems.

experience

Where he's worked.

2026 — present · Frankfurt
AI / ML Engineer
Deutsche Börse · CTO Area

Building agentic AI for capital-markets infrastructure: retrieval-augmented generation, multi-agent orchestration, and LLM-backed tooling, with a focus on reliability, evaluation, and guardrails in a regulated domain.

LangGraphRAGFastAPIGCP / Vertex AIKubernetes
2018 — 2025 · Stuttgart
Research / Software Engineer
Robert Bosch GmbH

Autonomous driving simulation and perception. Built synthetic data pipelines and Radar/Lidar sensor models, developed object detection and tracking, and applied neural rendering (NeRF, Gaussian Splatting) to sensor simulation. Later work extended into GenAI, RAG, and multi-agent systems. Output includes peer-reviewed publications and patents.

PerceptionNeRFGaussian SplattingSynthetic DataPyTorch
research

Topics he has covered include.

Neural rendering for sensor simulation
Synthetic data generation
Object detection & tracking
Radar / Lidar sensor models
projects

Things he's building.

Web App

Voyager

A self-hostable personal portfolio manager: broker CSV import, AI-powered natural-language queries over your holdings, German tax helpers, and a customizable widget dashboard.

Next.js · TypeScript · FastAPI · Docker
AI Tool

Read-it-Later, reimagined

An AI-powered read-it-later and content app: capture, summarize, and resurface long-form reading with LLM-assisted organization.

Python · LLMs · RAG
Hardware · Local LLM

Talking Action Figure

A DIY voice-interactive action figure running a local LLM on edge hardware. A build exploring on-device speech and personality.

Local LLM · Edge hardware · TTS / STT
Concept

h7ai

A portfolio and job-matching app built as a learning-by-doing project, pairing candidate profiles with AI-engineering roles.

Python · Embeddings · Matching
stack

Tools of the trade.

AI / ML

  • RAG & retrieval systems
  • LangGraph · multi-agent
  • Evaluation & guardrails
  • NeRF · Gaussian Splatting
  • PyTorch

Engineering

  • Python
  • FastAPI
  • TypeScript · Next.js
  • Docker · Kubernetes
  • GCP · Vertex AI

Domains

  • Autonomous driving
  • Perception (detection / tracking)
  • Synthetic data generation
  • Radar / Lidar sensor models
  • Agentic AI for finance

Let's talk.

Open to conversations about AI, perception, and applied ML. Email is the fastest way to reach him.