Looking for a job

Elvira Kushlak

@elvirakushlak

AI Operations Specialist at a US AI startup, designing benchmark datasets and evaluation guidelines for language and video models.

Italy

Message

What I'm looking for

I want to run evaluation and annotation projects end to end: the rubric, the dataset, the reviewers and the report.

At a US AI startup, I design benchmark datasets and evaluation questions from scratch to test language models’ reasoning and robustness. I assess production LLMs in real conversations and evaluate video-generation models for visual consistency, audio quality, prompt adherence and fidelity to reference images.

I write evaluation guidelines for other assessors, review their work and deliver structured feedback for model validation. I’m looking to run evaluation and annotation projects end to end, from the rubric and dataset to reviewers and the report.

At Movly, I built business cases for about 35 founders moving to France and Spain, and every application I prepared was approved. I also produced investor decks, business plans and financial models; I hold a Master’s in Digital Humanities and Digital Knowledge from the University of Bologna.

Experience

Work history, roles, and key accomplishments

Current

AI Operations Specialist

Current

Stealth AI Startup

Sep 2025 - Present (1 year 1 month)

Design benchmark datasets and evaluation questions to test LLM reasoning and robustness. Evaluate production LLMs and video-generation models, and write evaluation guidelines for other assessors.

LLM Evaluation Multimodal Evaluation Benchmark Design Rubric Design Evaluation Datasets Guideline Based Assessment Quality Review Data Labeling

Research Intern

CINEMATEK – Royal Film Archive of Belgium

May 2025 - Sep 2025 (4 months)

Researched pre-cinema objects in the archive's Wunderkammer collection and turned them into a research-through-design exhibition concept. Specified a light digital layer for the exhibition and documented the visitor journey as an interactive Twine prototype.

Research Exhibition Design Digital Layer Specification Sensors Projection Raspberry Pi Twine

Startup Analyst

Movly

Jun 2022 - Apr 2025 (2 years 10 months)

Built business cases for about 35 founders entering France and Spain, with a 100% approval rate. Produced around 50 investor decks, business plans, and financial models, and managed each client's timeline end to end.

Business Case Development Market Entry Investor Decks Business Planning Financial Modeling Project Management Competitor Research

International Logistics Manager

Boris Riger Decoration Atelier

Oct 2021 - Aug 2023 (1 year 10 months)

Ran cross-border procurement with suppliers in several countries and resolved recurring delivery delays. Built purchase-tracking tools and managed the director's schedule, meetings, and logistics.

Cross border Procurement Supplier Management Purchase Order Tracking Scheduling Social Media Management

Research Assistant

Higher School of Economics

Dec 2020 - Apr 2021 (4 months)

Managed digital archives and transcribed research interviews for a gender-studies research project.

Digital Archives Transcription Research Support

Education

Degrees, certifications, and relevant coursework

University of Bologna

Master of Science, Digital Humanities and Digital Knowledge

2023 -

Pursuing a Master's in Digital Humanities and Digital Knowledge, with coursework in linked open data, knowledge representation, data science, and software engineering for cultural heritage.