Machine learning engineer & researcher

Veli Ates.

From complex data
to systems you can inspect.

I build reproducible ML pipelines and research tools for visual data analysis, simulation, and evidence-grounded AI.

Explore my work View CV

DFKIMaster’s thesis research

CHIRA 2026DiversityLens paper accepted

1.6M+ framesAudited with DiversityLens

Selected work / 01–04

Research, made tangible.

All projects

01 / Featured researchCHIRA 2026 · Accepted

DiversityLens

Making the hidden structure
of visual datasets visible.

An automated demographic auditing pipeline developed during my master’s thesis at DFKI. Built to examine image and video collections at a scale manual inspection cannot reach.

My work: Dataset discovery, sampling, attribute inference, and reporting.
Scope: 1.6M+ frames across 10+ benchmark datasets.

Read the case study Code ↗

DiversityLens visualization of demographic distributions across visual datasets — Visual dataset auditingPython / Computer vision

02 / Independent explorationIn development

Cell Engine

Exploring how a cell
becomes a system.

A hepatocyte simulation project connecting biochemical state, stochastic processes, and interactive 3D anatomy. An ongoing exploration of research software and complex systems.

Focus: Inspectable mechanics, parameter provenance, and simulation tooling.
Boundary: Biological validation remains incomplete. Visual behavior and quantitative evidence are distinguished in the demo.

Explore the project Launch demo ↗

Cell Engine interface showing a 3D hepatocyte and biochemical state panels — Hepatocyte simulationPython / Interactive 3D

03 / Retrieval & evidence

AI Evidence Assistant

A document question-answering interface with source passages. Explore a guided sample of the retrieval workflow.

Open sample demo ↗Static demonstration

04 / ML implementation

Edge AI

Manual neural-network training and tensor export toward a C++ inference path. A look inside the steps behind a model.

Explore the pipeline ↗Work in progress

Background & direction

Curiosity, with
an engineering practice.

My background is in embedded systems, with master’s thesis research at DFKI. I’m interested in the point where machine learning, scientific questions, and reliable software meet.

I care about what a system can demonstrate: the data behind a result, the assumptions inside a model, and whether someone else can reproduce the work.

Let’s work on a good problem.

Open to AI/ML engineering roles, research collaborations, and PhD opportunities.

veli58ates@gmail.com